Browse the Radio 4 episodes

Series 1 (Nov - Dec 2001)
Episode 1 (13/11/2001)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/1654125.stm

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Episode 2 (20/11/2001)

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Episode 3 (27/11/2001)

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Episode 4 (04/12/2001)

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Episode 5 (11/12/2001)

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Episode 6 (18/12/2001)

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Series 2 (Nov - Dec 2002)
Episode 1 (12/11/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2411647.stm

Oral learning of mathematics: Learning mathematics by oral learning as opposed to traditional formal methods can be advantageous. Examples of this are discussed including street children who can do mathematics when selling goods but cannot do it in schools.

Episode 2 (19/11/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2474983.stm

Mathematics of skimming stones: The mathematics behind skimming stones across water is discussed.

Risk of getting breast cancer for women who drink: A study reported that women who drank alcohol had an increased risk of breast cancer. The level of this increased risk is discussed including absolute and relative risks.

Episode 3 (26/11/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2503929.stm

Episode 4 (03/12/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2524909.stm

Episode 5 (10/12/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2546353.stm

Episode 6 (17/12/2002)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2569485.stm

Bell-ringing and permutations: The mathematics behind peals of bells is discussed including permutations.

School league tables: Discussion of the history of school league tables and attempts to include a value-added element. See also 19/01/2006, 11/12/2006, http://www.bbc.co.uk/radio4/factual/learningcurve_20080616.shtml.

Series 3 (Feb - Mar 2003)
Episode 1 (18/02/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2754827.stm

Observing and recording animal and human behaviour: The ways in which animal and human behaviour can be observed and recorded is discussed.

Episode 2 (25/02/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2788355.stm

Media reporting of risk: How the media report the risk of transportation and other issues is discussed, looking at the balance between their need for a good story and responsible reporting of the actual risks.

Episode 3 (04/03/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2807335.stm

Claims of undercounting in the 2001 census: The results from the 2001 census showed large changes in population for some cities. It is suggested that this may be a result of undercounting, but it may also be the result of past overestimates.

Episode 4 (11/03/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2825791.stm

<font size="+1">Episode 5 (18/03/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2847443.stm

<font size="+1">Episode 6 (25/03/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2867745.stm

Number magic and coincidences: How numbers can sometimes appear to do magic things and how coincidences can just be mathematically bound to happen.

Series 4 (Jun - Jul 2003)
<font size="+1">Episode 1 (12/06/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2960160.stm

Location of audio file on BBC website unknown (if anyone can find it, please update this).

<font size="+1">Episode 2 (19/06/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/2987566.stm

Estimating fish stocks: Scientists and fishermen often disagree about the size of fish stocks. The sampling techniques of the scientists are criticised by fishermen who misunderstand the distinction between sampling and carrying out a census. The modelling of fish stocks is also discussed.

Press releases, absolute, relative risk, reporting results: Press releases about the results of cancer studies are condemned for stating only relative risk when absolute risk is vital to a proper understanding of the issues involved. A detailed explanation of absolute and relative risk is given. The broadcast goes on to consider publications that result from the work and how they are often worded to make the work sound more important and attract further funding.

<font size="+1">Episode 3 (26/06/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3013080.stm

Location of audio file on BBC website unknown (if anyone can find it, please update this).

<font size="+1">Episode 4 (03/07/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3026120.stm

<font size="+1">Episode 5 (10/07/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3044354.stm

GM crop trials, hypotheses and power: The nature of appropriate null hypotheses for trials of GM crops are discussed. The power of statistical tests is also discussed.

<font size="+1">Episode 6 (17/07/2003)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3065959.stm

Series 5 (Jan - Feb 2004)
<font size="+1">Episode 1 (08/01/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3381919.stm

Due dates for giving birth: Are doctors giving expectant mothers the correct due date for the birth of their baby? In France, they give longer than in the UK. A discussion of how the date is calculated is given including whether the mean length of gestation is a good measure or should the mode be used.

<font size="+1">Episode 2 (15/01/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3332015.stm

<font size="+1">Episode 3 (22/01/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3412909.stm

<font size="+1">Episode 4 (29/01/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3439027.stm

Speed cameras and regression to the mean: Effectiveness of speed cameras and regression to the mean A government report into the effectiveness of speed cameras is examined in relation to the issue of regression to the mean. See also 22/12/2005, 19/11/2007.

<font size="+1">Episode 5 (05/02/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3455985.stm

<font size="+1">Episode 6 (12/02/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3481431.stm

Series 6 (Jun - Jul 2004)
<font size="+1">Episode 1 (17/06/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3811679.stm

<font size="+1">Episode 2 (24/06/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3833043.stm

Confusion over incidence and prevalence of cancer: A study was reported to be saying that many more people were going to get cancer in the future. What it actually said was that in the future there would be many more people alive who had had cancer. The former is the incidence and the latter is the prevalence. This distinction is discussed and also the implications of the increase in prevalence.

<font size="+1">Episode 3 (01/07/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3852909.stm

<font size="+1">Episode 4 (08/07/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3874809.stm

Measuring impartiality: How one might go about measuring impartiality of media reports.

Relationships and causation: Various examples of how finding a relationship does not mean that there is a causal relationship that exists.

<font size="+1">Episode 5 (15/07/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3895797.stm

Childcare and antisocial behaviour: A report suggested that there was an association between children exhibiting antisocial behaviour and them being in childcare for many hours a week when they were very young. The size of the effect, the relevant comparison population (all children or children who had little childcare) and other relevant factors are discussed.

<font size="+1">Episode 6 (22/07/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3914227.stm

<font size="+1">Episode 7 (29/07/2004)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/3933513.stm

Location of audio file on BBC website unknown (if anyone can find it, please update this).

Series 7 (Jan - Feb 2005)
<font size="+1">Episode 1 (13/01/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4164055.stm

Risk of ear tumours from mobile phone use: A study suggested that the risk of getting an ear tumour was doubled for a certain category of mobile phone user. The usefulness of this relative risk without reference to the absolute risk is discussed.

When the median is more useful than the mean: Examples of when the mean is not a useful measure and it would be better to use the median.

<font size="+1">Episode 2 (20/01/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4187395.stm

Life expectancy after surgery is right-skewed: After cancer surgery, the median life expectancy was not very high, but further investigation revealed a right-skewed distribution which gave the hope of many further years of life.

<font size="+1">Episode 3 (27/01/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4205971.stm

International Passenger Survey: How the International Passenger Survey is carried out. This is a survey of travellers into and out of the U.K. and is used for various purposes including migration statistics and balance of payments. See also the story "Measuring Migration" broadcast on Radio 4 on 29/04/2016 (Radio 4 Series 37 Episode 5) and as part of the special series The Referendum by Numbers broadcast on 14/06/2014 (The Referendum by Numbers, episode 2: Immigration).

Measuring youth crime and overstating the level: A survey of youth crime got big headlines in the media. However, closer investigation of the data collection methods suggests that the levels of crime are not as great as suggested.

Simpson's Paradox: An example of Simpson's Paradox where averaging averages is not the same as calculating an overall average. See also Radio 4 Series 37 Episode 5 (29/04/2016), World Service Episode 225 (29/04/2016), Radio 4 Series 41 Episode 1 (25/08/2017).

<font size="+1">Episode 4 (03/02/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4230221.stm

Benford's Law: Benford's Law concerns the natural lack or randomness in apparently random data. An example given is that lengths of rivers, whatever measurement units you use, more often begin with the digit 1 than the digit 2. An example of how Benford's Law is used to detect fraud is also discussed. See also Benford's paper on the subject: Benford, F. (1938) "The law of anomalous numbers", Proceedings of the American Philosophical Society, 78, 4, pp551-572.

Counting the number of Somalis in the UK: Knowing the number of Somalis in the UK is important because some grants are awarded to minority groups on the basis of how many people they are serving. However, there is a large disparity between the numbers claimed and those counted in the 2001 census. The reasons for this are discussed, as are the various ways of coming up with an estimate.

<font size="+1">Episode 5 (10/02/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4249239.stm

<font size="+1">Episode 6 (17/02/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4263679.stm

Publication bias and its consequences: A discussion of the phenomenon of publication bias and the implications it can have: e.g. using inappropriate drugs, implementing inappropriate policies. See also 25/05/2012.

Series 8 (Jun - Jul 2005)
<font size="+1">Episode 1 (23/06/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4116908.stm

Animated graphics and conditions in the developing world: Various measureable items in the developing world have been changing over the years. This article describes how animated statistical graphics can be used to see how this is happening.

<font size="+1">Episode 2 (30/06/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4635239.stm

Passive smoking risks, CIs and reporting risks: The statistics behind the risks of passive smoking, a good discussion about what confidence intervals are and what they mean, and a discussion about the reporting of increased risks.

<font size="+1">Episode 3 (07/07/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4657179.stm

<font size="+1">Episode 4 (14/07/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4679743.stm

<font size="+1">Episode 5 (21/07/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4702683.stm

Taking into account more than just the average: Discusses the perils of just looking at the average when making decisions as the various possible outcomes may have an asymmetric distribution. With the use of examples, it is explained that the full range of outcomes need to be looked at.

Series 9 (Dec 2005 - Feb 2006)
<font size="+1">Episode 1 (22/12/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4541390.stm

Speed cameras and regression to the mean: Speed cameras and saving lives Do speed cameras really save lives? This is used as an example of the regression to the mean phenomenon. See also 29/01/2004, 19/11/2007.

<font size="+1">Episode 2 (29/12/2005)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4566512.stm

Choosing kidney transplant recipients: An explanation of how the decision is made concerning who should receive a kidney that is available for transplant, using various factors associated with th kidney available and those on the waiting list.

<font size="+1">Episode 3 (05/01/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4584962.stm

Data quality and related issues in the health services: Discussion of data quality in the health services and how it can be very poor and thus mean that only large differences between items of interest can be identified. Also mentioned are missing values and the cost of obtaining quality data.

<font size="+1">Episode 4 (12/01/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4604992.stm

Difficulties of constructing international league tables: The inherent difficulties of constructing league tables of nation states are discussed, including differences in measuring systems and differences in what is considered important to have measured.

<font size="+1">Episode 5 (19/01/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4627552.stm

Chances of having a boy or girl baby: Examines the probability of having a boy or a girl baby and whether having had several boys or several girls previously affects this.

Do left-handed people die younger: Examines the evidence surrounding the suggestion that left-handed people die younger, including analysis of data about sports people.

School league tables: Discusses Key Stage 3 to GCSE value-added tables. Because the top band at Key Stage 3 covered 20% of pupils, selective schools who take the top of this 20% seem to have better value-added that those schools left with the rest of this 20% because they are all put at the same starting point, when in fact the actual improvement for the top of the 20% might be less than the actual improvement for the rest. See also 17/12/2002, 11/12/2006, http://www.bbc.co.uk/radio4/factual/learningcurve_20080616.shtml.

<font size="+1">Episode 6 (26/01/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4646222.stm

Counting wild animals using roadkill and droppings: Talks to people who estimate the number of hedgehogs in the wild by counting the number of dead ones seen on roads and to people who count deer droppings to estimate the number of deer.

<font size="+1">Episode 7 (02/02/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/4672938.stm

Series 10 (Jun - Jul 2006)
<font size="+1">Episode 1 (22/06/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5075070.stm

<font size="+1">Episode 2 (29/06/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5127920.stm

<font size="+1">Episode 3 (06/07/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5153128.stm

First-borns and intelligence: Anecdotal evidence might suggest that first-born children end up more intelligent because of the extra time parents have to concentrate on them. A Dutch Army study backed this up but another study looking within families did not. See also Belmont. L. & Marolla, F.A. (1973) "Birth order, family size and intelligence", Science, 182, 4117, pp1096-1101 and Rodgers, J.L., Cleveland, H.H., van den Ord, E. & Rowe, D.C. (2000) "Resolving the debate over birth order, family size and intelligence", American Psychologist, 55, 6, pp599-612.

<font size="+1">Episode 4 (13/07/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5175432.stm

<font size="+1">Episode 5 (20/07/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5198274.stm

<font size="+1">Episode 6 (27/07/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/5215842.stm

Bad headline grabbing surveys: Examples and discussion of poor surveys that give headline grabbing conclusions.

Randomised control trials: What are randomised control trials and why are they so useful?

Series 11 (Oct - Dec 2006)
<font size="+1">Episode 1 (23/10/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6076738.stm

Diet and behaviour in prisons: Details of a careful study linking diet and behaviour in prisons. Ethical considerations are discussed, how the study was conducted. Questions are raised about the lack of follow-up studies and how the Home Office undertakes studies in general.

Hawthorne effect: How collecting data from people can change their responses. See also Radio 4 Series 29 Episode 5 (27/09/2013).

<font size="+1">Episode 2 (30/10/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6098740.stm

Average call time for call centre staff: Story of call centre staff who are judged on their average call duration. Some were reducing their average call duration by cutting people off after a couple of seconds.

Targets and payments in the NHS: Discusses gaming where the existence of targets and payments systems in the NHS may be affecting the recording of data and how hospitals carry out their activities.

<font size="+1">Episode 3 (06/11/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6120744.stm

Making performance measures fit a distribution: Managers had to rate the performance of their staff and were expected to have a range of good, average and poor staff. In order to achieve the required distribution of performances, managers were actively trying to get poor staff on their team to make the numbers work out.

<font size="+1">Episode 4 (13/11/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6142952.stm

The curse of Sports Illustrated: People appearing on the cover of Sports Illustrated appear to be prone to having a downturn in fortunes after being on the cover. This is an example of regression to the mean which is discussed.

<font size="+1">Episode 5 (20/11/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6165950.stm

Generating random numbers: Discussion of how random numbers are generated.

<font size="+1">Episode 6 (27/11/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6187896.stm

Average graduate salaries: A survey of graduates showed that geographers had the highest average wage. Further investigation showed the data included the basketball player Michael Jordan, showing the perils of looking just at the average. See also Radio 4 Series 42 Episode 1 (12/01/2018).

Randomness of V1 bombings: In WW2, it was a subject for debate whether the V1 rockets dropped on London were targeting various places or were dropping at random. A statistical analysis was carried out and concluded they were random.

<font size="+1">Episode 7 (04/12/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6205730.stm

<font size="+1">Episode 8 (11/12/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6169603.stm

School league tables: Discussion of the history of school league tables and attempts to include a value-added element. See also 17/12/2002, 19/01/2006, http://www.bbc.co.uk/radio4/factual/learningcurve_20080616.shtml.

<font size="+1">Episode 9 (18/12/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6189369.stm

2011 Census - issues raised and plans: Discussion of issues raised by the 2001 Census and plans for the 2011 Census.

Predicting hits in the music industry: Commercial software that analyses music and based on its characteristics, predicts whether or not it would a hit. It appears to use some sort of cluster analysis.

<font size="+1">Episode 10 (25/12/2006)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6197607.stm

Series 12 (Apr - Jun 2007)
<font size="+1">Episode 1 (23/04/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6575057.stm

<font size="+1">Episode 2 (30/04/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6607733.stm

<font size="+1">Episode 3 (07/05/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6626111.stm

<font size="+1">Episode 4 (14/05/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6653057.stm

<font size="+1">Episode 5 (21/05/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6675909.stm

High mortality rate is a statistical blip: A high mortality rate was observed for a Primary Care Trust. Why was this? Investigations revealed that it was most probably just a blip caused by a small sample size.

UNICEF survey of child welfare: Creating rankings from difference scores and conducting international surveys.

<font size="+1">Episode 6 (28/05/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6692577.stm

Social class and life expectancy: The whole programme, devoted to the evidence of a link between soical class and life expectancy, and how that evidence is obtained.

<font size="+1">Episode 7 (04/06/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6718007.stm

Crime statistics and public perception of levels of crime: Discussion of how official crime statistics show falling levels whilst the public perception is of increasing levels and how the presentation of the statistics is important in this context.

<font size="+1">Episode 8 (11/06/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/6741547.stm

Decision making and software to aid decision making: Discussion of ideas behind the decision making process and of software that helps with decision making.

Series 13 (Oct - Dec 2007)
<font size="+1">Episode 1 (29/10/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7063633.stm

<font size="+1">Episode 2 (05/11/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7078555.stm

<font size="+1">Episode 3 (12/11/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7090524.stm

Happiness and causality: Can we disentangle the causes of happiness from other factors? Talks about the need for panel data.

<font size="+1">Episode 4 (19/11/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7101633.stm

Counting and mapping birds: How the British Trust for Ornithology (BTO) count and map birds.

Measuring waiting times in the NHS: The different ways in which waiting times are measured in the NHS is discussed.

Speed cameras and regression to the mean: Speed cameras and reducing accidents The issue of whether speed cameras reduce accidents or not is revisited. It includes an explanation of regression to the mean. See also 29/01/2004, 22/12/2005.

<font size="+1">Episode 5 (26/11/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7112919.stm

<font size="+1">Episode 6 (03/12/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7124758.stm

Living on a dollar a day: The number of people living on less than a dollar a day is a well- known measure of poverty. This article discusses how a dollar a day is measured.

National positions in league tables: England and Scotland's positions in a league table of reading abilities are discussed with reference to the reliability of the information that goes into the table and the relevance of changes in positions over time when you have new countries taking part. See also Radio 4 Series 13 Episdode 7 (10/12/2007), World Service Episode 149 (15/11/2014) and Radio 4 Series 33 Episode 2 (09/01/2015).

Reporting risks related to diet and cancer: Should we talk about absolute risks as well as relative risk and about predicted benefits (e.g. in terms of months/years extra life expectancy) of changing diet?

<font size="+1">Episode 7 (10/12/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7136351.stm

National positions in league tables: National positions in a league table of mathematics abilities are given the same investigation as that of reading abilities. See also Radio 4 Series 13 Episode 6 (03/12/2007), World Service Episode 149 (15/11/2014) and Radio 4 Series 33 Episode 2 (09/01/2015).

Pregnancy-related deaths and weight: A report apparently linked the chances of dying due to pregnancy or childbirth and mother's weight. In fact it was poor reporting of the results that caused the headlines.

Survey of priorities for carers: A survey of pririties for carers reported that the main concern was for recognition from health-care workers rather than about money. An investigation showed that concern about money had been split between age-groups. When groups were combined, money was the main concern.

<font size="+1">Episode 8 (17/12/2007)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7148322.stm

Series 14 (Apr - May 2008)
<font size="+1">Episode 1 (07/04/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7324125.stm

<font size="+1">Episode 2 (14/04/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7346500.stm

<font size="+1">Episode 3 (21/04/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7358928.stm

<font size="+1">Episode 4 (28/04/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7370557.stm

Food influences baby gender: Explanation of p-values and adjusting for multiple testing. Mathews, F., Johnson, P.J. & Neil, A. (2008) You are what your mother eats: evidence for maternal preconception diet influencing foetal sex in humans, Proceedings of the Royal Society B: Biological Sciences.

Has global warming peaked: How looking at graphs of data in different ways may lead to different conclusions.

<font size="+1">Episode 5 (05/05/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7380853.stm

RICS survey of surveyors: The RICS Housing Market Survey of surveyors reports, for instance 78.5% more Chartered Surveyors reported a fall than a rise in house prices rather than tell us what percentage said a rise, what percentage said a fall and what percentage said no change. See also http://www.rics.org/Practiceareas/Property/Residential/Market/.

Whitehall II: Discusses a health study in the context of confounding. It also mentions factors that are predictive and then talks about them as if this means they are causal. See also 26/05/2008, http://en.wikipedia.org/wiki/Whitehall_Study, http://www.ucl.ac.uk/whitehallII/.

<font size="+1">Episode 6 (12/05/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7396017.stm

Telephone polls versus Internet polls: YouGov internet polls versus telephone polls - which method is better and how come YouGov got the London Mayor election correct?

<font size="+1">Episode 7 (19/05/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7408337.stm

<font size="+1">Episode 8 (26/05/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7417216.stm

Whitehall II: Follows on from discussion of Whitehall II study on 05/05/2008 and discusses causal narratives and that just because a statistical association between things has been found and that there is a plausible causal narrative, it still does not mean that a true association has necessarily been found. See also 05/05/2008, http://en.wikipedia.org/wiki/Whitehall_Study, http://www.ucl.ac.uk/whitehallII/.

Series 15 (Dec 2008 - Jan 2009)
<font size="+1">Episode 1 (05/12/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7765581.stm

Halo effect: In surveys asking people's opinions of various aspects of some organisation or person (e.g. aspects of business for firms), there is a halo effect where firms that score well in one area also score well in other areas, not because they are genuinely good in these other areas but because the people doing the scoring are inclined to favour them because they know they are good. There is also the example of George Bush's approval ratings going up after the September 11th attacks not just for foreign policy but also for economic policy. See also http://news.bbc.co.uk/1/hi/business/7758713.stm.

<font size="+1">Episode 2 (12/12/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7778304.stm

House numbers and the number 13: Some parts of the country do not use the number 13 when numbering houses in a street. This story talks about how the house numbering system came about and why some places use number 13 and some do not.

LIBOR: Looking at how the LIBOR is calculated and the possible effect that publicising the numbers that are used to calculate it could have on its trustworthiness.

<font size="+1">Episode 3 (19/12/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7790396.stm

<font size="+1">Episode 4 (26/12/2008)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7789699.stm

<font size="+1">Episode 5 (02/01/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7798152.stm

Average wages: How to calculate the average wage when the distribution of wages is skewed. The mean is not appropriate but the median is better. Which is used in the media is often not made clear.

Lake Wobegon effect: The Lake Wobegon effect occurs where the majority of people rate themselves as above/below average at something, like driving ability or politeness. This story investigates this effect and how it makes it difficult to get sensible data.

<font size="+1">Episode 6 (09/01/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7819674.stm

<font size="+1">Episode 7 (16/01/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7833271.stm

<font size="+1">Episode 8 (23/01/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/7845787.stm

Gossett, Fisher and significance: What is meant by statistical significance is discussed, referring back to the work by WS Gossett (alias Student) and Ronald Fisher. An example is given which appears to be wrong. An average of 20 and variance of plus or minus 10 is mentioned for a fictitious diet pill which is said to lead to a likely range of 10 to 30 pounds weight loss. A competing pill has an average weight loss of 5 but will vary by only 0.5 pounds on either side leading to a likely rage of 4.5 to 5.5. It is then claimed that doctors/dieticians/etc. would prefer the second pill because it is more precise and therefore significant. I suspect that the variance is really the standard error of the mean and that the intervals should thus be 0 to 40 for the first pill and 4 to 6 for the second pill. The second pill would then definitely be preferable.

Series 16 (Apr - May 2009)
<font size="+1">Episode 1 (17/04/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8003979.stm

<font size="+1">Episode 2 (24/04/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8016366.stm

<font size="+1">Episode 3 (01/05/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8028742.stm

Modelling the Swine Flu epidemic: In the early days of the Swine Flu epidemic, the programme talks to some people trying to model the spread of the virus. They talk about the sort of information that they need, the fact that they need to know the distribution of things like incubation times, etc. They also discuss the reporting of the story in the media and what sort of information journalists use.

Over-simplicity of finance equations: The use of formulae by financial institutions in making business decisions is discussed. Particular focus is put upon a formula which, in order to be simple enough to be applied quickly, ignores the fact that investment/lending opportunities are not independent of each other and their expected returns may be correlated.

<font size="+1">Episode 4 (08/05/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8040011.stm

Alcohol consumption and risk of cancer: A study reporting an increased risk of breast cancer in women that drank alcohol is examined. It is criticised for excluding from reports women who did not drink at all and got cancer. It is also pointed out that other studies show moderate alcohol consumption can be protective in the case of some cancers. The level of increased supposed risk is discussed. It is suggested that if a study with a million women is required to detect an effect then that effect is so small as to be of questionable relevance.

Understanding questions: Being familiar with the context of a question can hinder a correct response because the respondent may read too deeply into the question or have enough information to doubt their initial reaction. Also different cultures will understand questions in different ways.

<font size="+1">Episode 5 (15/05/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8051629.stm

Deal Or No Deal: The television programme Deal Or No Deal is discussed along with perceptions of luck and regret.

Digits of pi and normal numbers: A claim that as the expansion of pi cannot be written exactly in decimal form, any combination of digits must occur somewhere in the expansion is explained to be wrong. The idea of normal numbers is introduced (those where every two-digit number occurs on average one in every one hundred times in an expansion and every three-digit number occurs on average one in every one thousand times in an expansion, etc.). It is pointed out that a truly random number must be a normal number.

Statistics of domestic violence: The story of how a statistic about the levels of domestic violence transformed itself over time in the hands of the media into a completely incorrect and dramatic statement. Also about how statistics on domestic violence are and perhaps should be collected.

Who is hardest hit by the recession: Media reports are discussed which talked about the places/occupations that had the largest percentage increase in benefit claimants. Further investigation shows that some of these places/occupations have in fact had very small increases, but as they are from a small base, the percentage increase is large. It is argued that in this context, the change in percentage claiming benefits would be better than the percentage change in claimants.

<font size="+1">Episode 6 (22/05/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8061968.stm

MPs' expenses scandal and safe seats: An MP had noticed a relationship where the larger the majority that an MP had at the 2005 general election, the more likely they were to be named by The Daily Telegraph when it was reporting the MPs' expenses scandal. He divided the constituency majorities into quartiles and counted the number of MPs named by the paper in each group, showing a relationship. More detailed statistical analysis was suggested including a t-test. Comments on the MP's blog suggested other tests. See also http://markreckons.blogspot.com/.

Predicting football results: Chance plays a part in football results so it could be said that the team that finishes the season at the top of the table is not always the best team. Using simple modelling, one can say which teams are luckier or less lucky in a season. Also by doing some simple modelling, one can make predictions of football scores. See also 14/08/2009.

Series 17 (Aug - Sep 2009)
<font size="+1">Episode 1 (07/08/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8189480.stm

<font size="+1">Episode 2 (14/08/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8201615.stm

Predicting football results: A piece following-up from 22/05/2009 reports on the results of the matches predicted. See also 22/05/2009.

<font size="+1">Episode 3 (21/08/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8207622.stm

<font size="+1">Episode 4 (28/08/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8213670.stm

Conviction statistics for rape: How conviction statistics for rape are typically reported (convictions per complaint to police) compared with other crimes (typically convictions per court trial). Also how comparisons with other crimes and countries are made difficult by collection and recording methods.

<font size="+1">Episode 5 (04/09/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8234117.stm

<font size="+1">Episode 6 (11/09/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8248922.stm

Attractive people have more girl babies: An analysis of data apparently showed that attractive people were more likely to have girl babies than unattractive people. The author of the analysis had also found unlikely sex ratio differences for other categories of people. The analyses were challenged by Andrew Gelman who suggested that they were just statistical anomalies due to random fluctuations in small samples, particularly as the analyses reported appeared to be chosen because of their significance. The excerpt discusses regression, statistical significance, using an ordered category as if it were continuous, combining ordered categories to produce binary variables. It also discusses sample sizes and effect sizes in surveys.

Series 18 (Dec 2009 - Jan 2010)
<font size="+1">Episode 1 (11/12/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8404720.stm

Reliability of wine ratings: A discussion of how wine competitions are judged and medals awarded with particular reference to how reliable the ratings are and how reproducible the judgements might be.

<font size="+1">Episode 2 (18/12/2009)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8418233.stm

Surveys of water voles: A 2009 survey of water vole sightings suggested that there were more around than when a 2008 survey was carried out. However, the methodology of the survey had changed so little could be said to compare 2008 with 2009.

<font size="+1">Episode 3 (01/01/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8424658.stm

<font size="+1">Episode 4 (08/01/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8448326.stm

Exponential growth: The exponential function is discussed with particular reference to exponential growth. Population growth and the idea of Malthus are discussed. A YouTube presentation is referenced.

Passenger profiling to combat the threat of terrorism: Discussion of the use of passenger profiling in order to determine who should be subjected to more thorough security checking. It is suggested that the very fact that passenger profiling is carried out changes the statistical basis for the profiling as terrorists change tactics to try and circumvent the profiling. The need to use random checks is explained and a scoring mechanism for determining who to check is suggested.

<font size="+1">Episode 5 (15/01/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8461211.stm

First past the post polling and conversion into Parliamentary seats: A discussion of how the overall share of the vote converts into Parliamentary seats in different ways for different political parties.

Using Wales as a unit of measurement: Often news reports in the U.K. use the size of Wales as a means of comparison when trying to explain how big some things are. This phenomenon is discussed.

<font size="+1">Episode 6 (22/01/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8474881.stm

Statistics and politics - from early days to the UK Statistics Authority: Since its early days, Statistics has had an uneasy relationship with politics and politicians. The story of this relationship is told, up to the introduction of the UK Statistics Authority.

Series 19 (May - Jun 2010)
<font size="+1">Episode 1 (21/05/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8697176.stm

Voting systems, voting power and the alternative vote system: With the 2010 General Election resulting in a coalition agreement that promised a referendum on changing the first past the post electoral system to the alternative vote system, this extract discusses different voting systems. It explains that no electoral system can be perfect. Also discussed is the idea of voting power that represents how much power a party holds in Parliament.

Why opinion polls failed to correctly predict 2010 General Election result: Opinion polls during the 2010 General Election campaign predicted that the Liberal Democrats would get a much larger share of the vote than they ended up getting in the election itself. The reasons for this are discussed and also why exit polls on election night got it right.

<font size="+1">Episode 2 (28/05/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8711249.stm

Standardised mortality ratios (SMRs) as measures of patient care: SMRs can be useful in detecting when things are going wrong with health care but they are reported in a misleading way by the media. They are not able to predict how many excess deaths occurred as a result of poor treatment. See also 01/10/2010.

<font size="+1">Episode 3 (04/06/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8720866.stm

Pareto Principle: The Pareto Principle or "80-20 rule" is discussed, giving examples of where it seems to work (e.g. 20% of the population have 80% of the wealth) and its history.

Ugliness and criminality: Research investigating attractiveness and the propensity to commit crime is discussed along with the wider topic of anthropometrics. The work of Quetelet and Galton are discussed.

<font size="+1">Episode 4 (11/06/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8734972.stm

Laffer curve: The Laffer curve describes a phenomenon where at low rates of taxation, more can be raised by increasing the tax rate but at higher tax rates, increasing the rate yields less money as tax avoidance strategies are used. The concerning raising the rate of capital gains tax is used as an example where the Laffer curve may be appropriate and the issue of where the peak of the curve might lie is discussed.

Scale for reporting earnings: In a similar way to which areas of land are often referred to as fractions of multiples of "the size of Wales", a listener suggests a scale for reporting earnings relative to the income of a pensioner.

Football statistics: The collection of statistics on many different aspects of games of football and performances of footballers is discussed with particular reference to the then forthcoming 2010 FIFA World Cup. Based on the data collected, it is suggested that a particular goalkeeper (Joe Hart) should be used and that England perform badly when Steven Gerrard and Frank Lampard play together. Although England's performance in the tournament suggest that there might be something in what the statistics suggest, the excerpt provides the opportunity to discuss the dangers of predicting beyond the circumstances of the data collection (Premiership performances predicting performance in World Cup) and of looking at headline statistics while not taking other circumstances into account (have England performed better with Frank Lampard not playing with Steven Gerrard because of the standard of opponents on those occasions?).

<font size="+1">Episode 5 (18/06/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8748159.stm

Measuring the gender pay gap: How one might go about measuring the gap between pay for men and women is discussed. A study into the reasons for the gap is presented which highlights the issue of career breaks affecting income.

Beautiful information: Listeners are asked to contact the programme to say what they were doing when listening so that David McCandless could create a graphical representation of the information. See also 25/06/2010.

Filling football sticker books: The number of stickers needed to complete football sticker books is discussed. A listener worked out how many one would expect to have to buy to get a complete set and the logic of his calculations are discussed. It is acknowledged that the calculations do not take account of swapping of stickers. What is not mentioned is that they have been assuming that the difference stickers are all produced in the same numbers and evenly distributed.

<font size="+1">Episode 6 (25/06/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8759208.stm

Beautiful information: Discusses some of the information sent in by listeners and the process undertaken by David McCandless to create the graphical representation (online representation not available on 03/09/2010). There is the opportunity to point out that much of what David McCandless is doing to create the representation is coding of the responses. See also 18/06/2010.

Equality of toilet facilities: Giving men and women equality in terms of toilet facilities is more complex than it might seem. On what basis do you measure equality? The number of facilities, the floor area devoted, the waiting time, etc.? The measurement issues involved in this matter are discussed.

Penalty shoot-out probabilities: Ahead of the England vs Germany game in the FIFA 2010 World Cup, the probability of England defeating Germany in a penalty shoot-out is calculated, based on past data and using the Binomial distribution, to be 1.4%. In the game itself, Germany won 4-1 with a goal by Frank Lampard (with the score at 2-1) being ruled out despite TV pictures showing it had cross the goal line.

Series 20 (Aug - Oct 2010)
<font size="+1">Episode 1 (27/08/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8949827.stm

Income inequality and social, health issues: A book looking at the link between income inequality and social/health issues is discussed. There is a discussion about how outliers can be influential in make it appear that a relationship exists whereas when they are removed, the relationship is much more questionable. Also the importance of carrying out multivariate modelling to investigate social phenomena rather than just bivariate scatterplots is discussed.

<font size="+1">Episode 2 (03/09/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8966509.stm

Calculating life expectancy: Various ways of calculating and expressing life expectancy are given. Also a discussion about how life expectancy may grow significantly in future years.

Relative safety of cycling, walking and driving: The relative safety of cycling, walking and driving is discussed. Whether it is better to look at it in terms of per billion km travelled or per hour travelled, etc. is discussed. Should we look at deaths or injuries?

<font size="+1">Episode 3 (10/09/2010)

URL: http://news.bbc.co.uk/1/hi/programmes/more_or_less/8987489.stm

Definitions of disability: Data from two different studies were put together for a newspaper article about discrimination against people with disabilities. It emerges that the definition of "disabled" was likely to be different in the two studies and serves as a reminder that in surveys one must take care in what one assumes respondents will have been answering.

<font size="+1">Episode 4 (17/09/2010)

URL: http://www.bbc.co.uk/programmes/b00tq1vk

<font size="+1">Episode 5 (24/09/2010)

URL: http://www.bbc.co.uk/programmes/b00tt6r6

<font size="+1">Episode 6 (01/10/2010)

URL: http://www.bbc.co.uk/programmes/b00tznbk

Needless deaths: David Spiegelhalter comments on the reporting of a study he lead concerning differing mortality rates in hospitals after children's heart surgery. The phrase "excess deaths" is criticised and a discussion of when mortality rates above average may or may not be a cause for concern takes place. See also 28/05/2010.

Series 21 (Dec 2010 - Jan 2011)
<font size="+1">Episode 1 (10/12/2010)

URL: http://www.bbc.co.uk/programmes/b00wdkf9

Modelling time spent at traffic lights: A London bus driver had been keeping records of how long he had to wait at traffic lights on his route and suggested that removing traffic lights might improve journey times. The way in which time spent at traffic lights is modelled is discussed.

<font size="+1">Episode 2 (17/12/2010)

URL: http://www.bbc.co.uk/programmes/b00wlh2t

Media misinterpretation of correlation as causality: How a story was invented and distributed to the media to see if they would misinterpret it as a causal link. It concerned genuine data about places with more mobile phone masts having higher birth rates. A discussion about the misinterpretation of correlation as causality is included.

<font size="+1">Episode 3 (24/12/2010)

URL: http://www.bbc.co.uk/programmes/b00wqgmq

<font size="+1">Episode 4 (31/12/2010)

URL: http://www.bbc.co.uk/programmes/b00wrbrx

<font size="+1">Episode 5 (07/01/2011)

URL: http://www.bbc.co.uk/programmes/b00x44l1

Proofiness: Charles Seife is author of "Proofiness" which concerns situations when numbers are used to deliberately confuse or make one side of an argument look better than it is. Ideas of false precision are included.

Psychic animals: Following the apparent success of an octopus in predicting results in the football World Cup in 2010, the programme asked a monkey to predict results in the Ashes test matches in 2010/11. In this excerpt they reveal that their apparently successful monkey was just one of many monkeys they "asked" and they use this to demonstrate that when we only hear about successful or unusual stories, we can get a false perception of a story. The continued success of the octopus in the World Cup, even after having been the one animal (out of many) to have gained press attention was unusual and perhaps suggested that the octopus might be psychic. Bayes logic is then used to combine a very low prior probability that one would have with the data to still conclude that the octopus is not psychic.

<font size="+1">Episode 6 (14/01/2011)

URL: http://www.bbc.co.uk/programmes/b00xb0sc

<font size="+1">Episode 7 (21/01/2011)

URL: http://www.bbc.co.uk/programmes/b00xj13v

Blue Monday: Several British newspapers and media outlets reported that a certain Monday in January 2011 had been calculated to be the most depressing day of the year. The programme looked into this research and was sent an equation by the researcher which explained how it was calculated. It turned out not to be based on any data at all and was constructed on the basis of the researcher's conversations with other people. More careful examination of the formula revealed inconsistencies such as it being impossible to calculate for a person if they had not made any New Year's resolutions and for people with substantial savings, the most depressing day in 2011 was calculated as being in 2010.

Series 22 (Apr - May 2011)
<font size="+1">Episode 1 (01/04/2011)

URL: http://www.bbc.co.uk/programmes/b00zt4h3

2011 Census - apparently ambiguous questions and the future of the census: Broadcast shortly after the 2011 Census, this discusses some apparently ambiguous questions on the Census and talks about possible plans for its future.

<font size="+1">Episode 2 (08/04/2011)

URL: http://www.bbc.co.uk/programmes/b0100j90

Recording, calculating and comparing youth unemployment figures: The difficulties of recording and calculating youth unemployment figures are discussed with particular reference to how students in part-time work are treated. The difficulties in making comparisons with past figures are highlighted.

Social mobility - a potential case of regression to the mean: Evidence for a lack of social mobility is examined. A study apparently shows that less able children from well off backgrounds catch up and overtake more able children from poorer backgrounds. But could this evidence just be a case of regression to the mean?

<font size="+1">Episode 3 (15/04/2011)

URL: http://www.bbc.co.uk/programmes/b0106vj6

Statistical significance - what it is and problems associated with it: Statistical significance is discussed and illustrated by drawing balls out of a bag. Issues associated with its application and how it does not incorporate prior information are discussed.

<font size="+1">Episode 4 (22/04/2011)

URL: http://www.bbc.co.uk/programmes/b010fd86

<font size="+1">Episode 5 (29/04/2011)

URL: http://www.bbc.co.uk/programmes/b010mwbt

<font size="+1">Episode 6 (08/05/2011)

URL: http://www.bbc.co.uk/programmes/b010xykh

Poverty line and measuring poverty: A common method of defining poverty is using a level relative to the median income in a country. However, this means that different countries have different poverty lines and a decrease in a country's median income can actually reduce the number of people living below the poverty line. This issue is discussed along with other ways of measuring relative and absolute poverty.

Busting and challenging stereotypes using statistics: Danny Dorling of Sheffield University gives some interesting and surprising statistics regarding the social make up of the U.K.

<font size="+1">Episode 7 (13/05/2011)

URL: http://www.bbc.co.uk/programmes/b010xzs5

Series 23 (Aug - Sep 2011)
<font size="+1">Episode 1 (05/08/2011)

URL: http://www.bbc.co.uk/programmes/b012x12m

IQs of internet browser users: A press release stated a survey had found that users of different internet browsers had different IQ levels and, in particular, that users of Internet Explorer had very low IQs. It subsequently turned out to be a hoax survey reported by a hoax company. However, before being revealed as a hoax, the story had been covered by many media outlets, including the BBC who had even spoken to experts who told them it was unlikely to be true. See also http://www.bbc.co.uk/news/technology-14370878, http://understandinguncertainty.org/spotting-hoax-using-statistics.

<font size="+1">Episode 2 (12/08/2011)

URL: http://www.bbc.co.uk/programmes/b0132pk7

<font size="+1">Episode 3 (19/08/2011)

URL: http://www.bbc.co.uk/programmes/b0138yld

Cochrane Collaboration review of salt in diets: A meta-analysis study by the Cochran Collaboration concluded that there was insufficient evidence to claim that reducing salt in diets would improve health outcomes. However, a summary of the work in non-scientific language used words that suggested that no benefit had been found from reducing salt intake, failing to clearly distinguish between non-significance due to lack of effect or lack of evidence. A discussion takes place about what studies should be included in the review process and p-values are explained (although see also 26/08/2011).

<font size="+1">Episode 4 (26/08/2011)

URL: http://www.bbc.co.uk/programmes/b013gd28

Confidence and p-values: In a story the previous week (19/08/2011 "Cochrane Collaboration review of salt in diets") p-values had been explained and it had been said that if you were "93% confident" then this would mean that statistical significance at the 5% level had not been achieved. This is technically wrong and a further discussion of this issue and explanation of p-values is given.

<font size="+1">Episode 5 (02/09/2011)

URL: http://www.bbc.co.uk/programmes/b013r2h4

Power of the placebo: The placebo effect is investigated with examples of studies involving acupuncture and spiritual healing. The importance of having a placebo included in studies is addressed. Studies that aim to investigate the placebo effect are discussed as is the potential usefulness of the effect in the National Health Service.

Seventy years of social surveys at the Office for National Statistics: A review of social surveys conducted by the Office for National Statistics (and its predecessors) between 1941 and 2011 revealing the changing social climate of Britain over 70 years, covering attitudes towards immigration, levels of drinking, smoking, etc. They speak to someone who has been doing face to face interviews for ONS surveys for 40 years, hearing of her experiences and her desire to fully undertake the interviews. This could easily lead on to a discussion as to why it is important to persist in trying to get interviews and avoid bias in non-responses.

<font size="+1">Episode 6 (09/09/2011)

URL: http://www.bbc.co.uk/programmes/b014641t

Series 24 (Dec 2011 - Jan 2012)
<font size="+1">Episode 1 (02/12/2011)

URL: http://www.bbc.co.uk/programmes/b017mz3x

Affordable housing - definitions: The importance of accurate, consistent definitions when discussing statistics. Also mentions the UK Statistics Authority asking for government statistics concerning this issue to be looked at.

The confusing world of percentages: Shares that suffer a 75% reduction in price and then a 200% increase sound like they end up at a higher level than at the start. This is not so and is illustrated with a story about Thomas Cook.

<font size="+1">Episode 2 (09/12/2011)

URL: http://www.bbc.co.uk/programmes/b017x76r

Supermarket prices: Almost all the supermarkets claim that they have low prices but how can they all be low? It turns out this is due to people buying different sorts of goods in different supermarkets and so it is perfectly possible for your shopping in Supermarket A to be cheaper than if you had bought the same products at Supermarket B and simultaneously the goods you would buy in Supermarket B be cheaper than if you bought those same goods in Supermarket A. A price check done by "The Grocer" does things differently.

Double yolks in eggs: The probability of finding eggs with double yolks is discussed and the chances of finding more than one "double yolker" in a single box of six eggs. This leads to a discussion of random and non-random events and independence of events. There is also mention of "short scale" billions and trillions as terminology for "American" billions and trillions as opposed to "British" billions and trillions. See also 16/12/2011.

FA Cup draw probabilities: After Manchester United were drawn to play Manchester City in the third round of the 2011/12 FA Cup there was much discussion about how unlikely this was to happen. It is explained that it was a 1 in 63 chance but also that the chances of "something interesting" happening in the draw would be more than that and that chances of the Manchester derby being drawn as the first fixture in the draw would be less than that.

<font size="+1">Episode 3 (16/12/2011)

URL: http://www.bbc.co.uk/programmes/b0184w5t

Probability and the Higgs boson: Scientists at CERN using the Large Hadron Collider to try and detect the Higgs boson reported initial findings in terms of there being "2 sigma evidence" for the existence of the particle. What they meant by this is that a traditional hypothesis test was significant at the 5% level. It is pointed out that this is not the same as there being just a 5% probability of the result being found by chance but that if the null hypothesis is correct then there is a 5% chance of seeing this (or more extreme) evidence. However, the discussion does not go on to say what the physicists should be saying or how they could calculate a probability that matches what they are saying. See also World Service Episode 26 (07/07/2012) and 28/12/2012.

Double yolks in eggs: More reasons why a whole box of six eggs might contain double yolks. See also 09/12/2011.

<font size="+1">Episode 4 (23/12/2011)

URL: http://www.bbc.co.uk/programmes/b018ft11

Measuring income inequality: How much more income do the wealthiest 1% in society have compared with the remaining 99%? This is not an easy question to answer because of differences in ways of measuring income between countries and even within countries. See also Radio 4 Series 38 Episode 2 (05/08/2016).

<font size="+1">Episode 5 (30/12/2011)

URL: http://www.bbc.co.uk/programmes/b018gzqx

Probability does not exist: David Spiegelhalter states that probability does not exist. He uses the example of a coin being tossed. The person tosses the coin can see the result and knows that the probability of "heads" is 100% but another person can only say "50%". The probability of a person with a particular set of characteristics having a heart attack in the next 10 years can be calculated but that is different from the probability of a particular individual having a heart attack in the next 10 years, even if that person matches the characteristics exactly.

<font size="+1">Episode 6 (06/01/2012)

URL: http://www.bbc.co.uk/programmes/b018xy2v

Bayes' Theorem and probability in the courtroom: The use of probability and Bayes' Theorem in court proceedings is discussed. The Prosecutor's Fallacy and Defendant's Fallacy are explained. It is emphasised that probabilities should not be regarded as being able to provide incontrovertible evidence of guilt/innocence but should be regarded as adding to the general body of evidence.

Calculating the percentage rise in train fares: When train fares rose at the beginning of 2012, the rise was expressed as a percentage increase. However, some fares rose by more than others, so what method was used to calculate the single figure quoted? Different possibilities for calculating this percentage are explored.

National Lottery probabilities: Is the probability of the numbers 1, 2, 3, 4, 5, 6 being drawn in the National Lottery the same as any other six numbers being drawn? Although it is referred to in a fictitious dramatisation. the issue of numbers being drawn in a particular order rather than in any order is not tackled. The idea of maximising expected winnings by choosing numbers that other people do not choose is also included. See also 13/01/2012.

<font size="+1">Episode 7 (13/01/2012)

URL: http://www.bbc.co.uk/programmes/b0196v3z

Executive pay: Different analyses reported different levels of increase in executive pay. The discussion highlights how trying to calculate the same thing, but with different types of sample, can lead to different results. Also discussed is the choice of mean or median as a measure of average and an attempted defence of using the mean. National Lottery probabilities: The issue of the dramatization in 06/01/2012 referring to numbers being drawn in a particular order is addressed. See also 06/01/2012.

Series 25 (Apr - May 2012)
<font size="+1">Episode 1 (20/04/2012)

URL: http://www.bbc.co.uk/programmes/b01g65h9

<font size="+1">Episode 2 (27/04/2012)

URL: http://www.bbc.co.uk/programmes/b01ghgsy

<font size="+1">Episode 3 (04/05/2012)

URL: http://www.bbc.co.uk/programmes/b01gvwy2

Lengths of border queues: Disagreements about how bad queues at Heathrow Airport had been lead to an investigation of what sampling methods were used to collect the data (sampling by time or by people). Also considered are issues of quoting the "worst" wait rather than average wait and what constitutes an official statistic.

Calculating probabilities of coincidences: A listener contacted the programme about a coincidence of people from the same location meeting elsewhere. A discussion of the probability of this happening is discussed along with issues of what data can be used for these purposes. See also World Service Episode 23 (16/06/2012).

<font size="+1">Episode 4 (11/05/2012)

URL: http://www.bbc.co.uk/programmes/b01h7cf2

<font size="+1">Episode 5 (18/05/2012)

URL: http://www.bbc.co.uk/programmes/b01hl4h2

Cat population growth: Suggestions that without a neutering programme, the population of cats could rise dramatically are investigated. As part of this a simulation study is undertaken in Excel, making simple assumptions about the underlying parameters. A discussion of making more realistic assumptions also takes place.

<font size="+1">Episode 6 (25/05/2012)

URL: http://www.bbc.co.uk/programmes/b01hxtmp

Publication bias and its consequences: Ben Goldacre talks about publication bias and its implications, particularly for randomised control trials (RCTs). Examples are given and efforts for overcoming the problem for RCTs are discussed. See also 17/02/2005.

Series 26 (Jul - Aug 2012)
<font size="+1">Episode 1 (13/07/2012)

URL: http://www.bbc.co.uk/programmes/b01kktmg

<font size="+1">Episode 2 (20/07/2012)

URL: http://www.bbc.co.uk/programmes/b01kt7d3

Trends and blips in road deaths: A set of data showed the number of road deaths increasing in one year after a declining trend. A discussion of how much weight to put on this figure takes place.

<font size="+1">Episode 3 (27/07/2012)

URL: http://www.bbc.co.uk/programmes/b01l1g64

Longevity of mothers of UK and US politicians: It has been observed that the mothers of US Presidential candidates seem to be longer lived than one would expect. A comparison with mothers of party leaders in UK General Elections is reported. The ideas of using means and medians are discussed along with Normal distribution curves. The data are compared with life expectancies in the different countries, taking into account the cohorts to which the mothers belong. See also World Service Episode 39 (06/10/2012).

<font size="+1">Episode 4 (03/08/2012)

URL: http://www.bbc.co.uk/programmes/b01l8rbq

<font size="+1">Episode 5 (10/08/2012)

URL: http://www.bbc.co.uk/programmes/b01ljwmx

<font size="+1">Episode 6 (17/08/2012)

URL: http://www.bbc.co.uk/programmes/b01lv7yn

Deaths and injuries in 20mph zones: A report that deaths and injuries have increased in 20mph zones is easily explained by there being more 20mph zones now than in past years. A discussion of the available evidence is given, along with a discussion of randomisation and existing studies of the true effect of 20mph zones.

RPI and the Carli Index: Details of how the RPI is calculated are given and it is suggested that due to its use of the Carli Index, it may consistently overestimate the level of inflation.

Series 27 (Nov 2012 - Jan 2013)
<font size="+1">Episode 1 (23/11/2012)

URL: http://www.bbc.co.uk/programmes/b01ny0fc

Counting casualties in the Israeli-Palestinian conflict: The number of casualties in the latest stage of the Israeli-Palestinian conflict was examined following claims made by one side. It emerged that it is hard to determine exactly what being a casualty means. If it is defined as being treated by medical staff then this includes people treated for shock but no physical injuries. The idea of it being easier to count deaths is discussed. See also World Service Episode 46 (24/11/2012).

Misreporting the number of cod in the North Sea: Newspapers were reporting that the number of cod in the North Sea were down to extremely small and implausible numbers. An explanation of how these reports came out is given along with information on how this sort of estimation is done. See also World Service Episode 38 (29/09/2012).

Fergie time: In English football's Premier League it is often alleged that Manchester United get generous amounts of injury time added on by the referee when they are chasing a goal to draw or win a match. This has been nicknamed "Fergie time"? after their manager Alex Ferguson. An examination of whether this is real or not is undertaken, including a discussion of how the examination is carried out. See also World Service Episode 46 (24/11/2012).

<font size="+1">Episode 2 (30/11/2012)

URL: http://www.bbc.co.uk/programmes/b01p0vqp

Wording of questions can make a difference: The polling company YouGov undertook surveys for three different organisations about the same issue (whether press regulation in the U.K. should be backed by law or not). The surveys revealed quite different results because of the way in which the questions had been asked. It was suggested that the questions had been deliberately worded in such a way that the client commissioning the survey got the answer they were looking for. An interview with Peter Kellner, President of YouGov took place and he denied this accusation saying that all the surveys were available on the web for people to see the differences. The piece also includes a classic recording from the television programme "Yes Minister"? about the order in which questions are asked influencing responses.

<font size="+1">Episode 3 (07/12/2012)

URL: http://www.bbc.co.uk/programmes/b01p4514

Morning sickness and the probability of having twins: Following the Duchess of Cambridge's hospitalisation with acute morning sickness, newspaper were speculating that she might be having twins. The association between acute morning sickness and having twins is investigated and the distinction between relative risk and absolute risk is discussed. See also World Service Episode 48 (08/12/2012).

<font size="+1">Episode 4 (14/12/2012)

URL: http://www.bbc.co.uk/programmes/b01p7hdn

Describing numbers with words: A discussion of how we use words to describe numbers, particularly in reference to events with small probabilities including "rare"? which is sometimes used in a very subjective way but also has a very particular meaning when applied to diseases. See also World Service Episode 49 (15/12/2012).

<font size="+1">Episode 5 (21/12/2012)

URL: http://www.bbc.co.uk/programmes/b01pcwqx

Delays in recording deaths: A discussion of how the timing of the registration of deaths is affected by the coroners and the impact this may have on accurate collection and reporting of death statistics.

Chocolate consumption and Nobel Prizes: A revised version of the World Service broadcast World Service Episode 45 (17/11/2012) which includes discussion of the Ecological Fallacy and the ordering of events when considering causation.

Identifying first-time house buyers: An investigation into how the average age of first-time house buyers was calculated. The difficulty in identifying first-time buyers and the data available to do this with is discussed.

<font size="+1">Episode 6 (28/12/2012)

URL: http://www.bbc.co.uk/programmes/b01pglrw

Probability and the Higgs boson: As one of the "Statistics of the Year"?, the way in which probability was used to make claims about the Higgs boson being found is discussed. See also Radio 4 Series 24 Episode 3 (16/12/2011) and World Service Episode 26 (07/07/2012).

<font size="+1">Episode 7 (04/01/2013)

URL: http://www.bbc.co.uk/programmes/b01phn0y

Series 28 (May - Jun 2013)
<font size="+1">Episode 1 (03/05/2013)

URL: http://www.bbc.co.uk/programmes/b01s4vf8

<font size="+1">Episode 2 (10/05/2013)

URL: http://www.bbc.co.uk/programmes/b01s8vvv

Suicide rates amongst ex-military personnel who served in wars: The number of ex-military personnel who served in particular wars who committed suicide is sometimes higher than the number of military personnel who died in the war, or the number of suicides seems very high. However, a better understanding of what is really going on is found by looking at the rates of suicide rather than absolute numbers and taking account of the populations in which the rates are calculated. See also Radio 4 Series 28 Episode 3 (17/05/2013).

<font size="+1">Episode 3 (17/05/2013)

URL: http://www.bbc.co.uk/programmes/b01sdw2d

Angelina Jolie's cancer risk: Angeline Jolie underwent a double mastectomy as a result of a probability calculation that gave a high chance of her developing breast cancer. The background to calculations of this type is given. See also World Service Episode 71 (18/05/2013).

Unreliability of surveys undertaken for PR reasons: Michael Gove, Education Secretary, stated that many surveys had shown that children's knowledge of historical facts was poor. It turns out that these surveys were conducted for public relations purposes rather than being academic research. The danger of placing much faith in surveys of this kind are discussed.

Counting jobs in the private and public sectors: The difficulty in counting jobs in the private and public sector is discussed. Public sector jobs which are outsourced to the private sector make it seem like the private sector is generating new jobs, changes in the definition of some sectors also make it seem changes are occurring as do schemes aimed at helping the unemployed into work.

Suicide numbers not interesting: The author of a report into suicide numbers in ex-military personnel was interviewed before the report was published but once the report was published, interviews for later in the same day were cancelled as the numbers were not as bad as many had been expecting. This story is linked to the issue of publication bias. See also "Suicide rates amongst ex-military personnel who served in wars". See also Radio 4 Series 28 Episode 2 (10/05/2013).

<font size="+1">Episode 4 (24/05/2013)

URL: http://www.bbc.co.uk/programmes/b01sjn5r

<font size="+1">Episode 5 (31/05/2013)

URL: http://www.bbc.co.uk/programmes/b01snyk3

<font size="+1">Episode 6 (07/06/2013)

URL: http://www.bbc.co.uk/programmes/b02142z5

Chances of having three sets of twins: The issues surrounding the calculation of the probability of having three sets of twins are discussed. Independence and the lack of it is included.

Statistical stories related to parenthood: The author of the book "Naked Statistics" gives some anecdotes concerning parenthood from his book. See also World Service Episode 74 (08/06/2013).

Men think about sex every seven seconds: It is said that men think about sex every seven seconds. This appears to be implausibly often and an American study investigates this. See also World Service Episode 74 (08/06/2013).

Series 29 (Aug - Oct 2013)
<font size="+1">Episode 1 (30/08/2013)

URL: http://www.bbc.co.uk/programmes/b038zhb7

Badger numbers: As a cull of badgers begins to try and control the spread of bovine TB, issues concerning the counting of badgers are discussed. Included is a 60% confidence interval which (conveniently?) gives a range of almost exactly 2,000 to 3,000.

Hospital Standardised Mortality Ratios (HSMRs): Hospital Standardised Mortality Ratios (HSMRs) are explained. It is emphasised that it is incorrect to regard these as identifying "excess" or "needless" deaths. See also Radio 4 Series 29 Episode 5 (27/09/2013).

Big data: A discussion of what is meant by "Big Data", its uses and anonymity issues. See also World Service Episode 63 (23/03/2013), Radio 4 Series 31 Episode 7 (13/06/2014) and World Service Episode 145 (17/10/2014).

<font size="+1">Episode 2 (06/09/2013)

URL: http://www.bbc.co.uk/programmes/b039dbk8

Selling houses to fund care home fees: A headline in the Daily Telegraph gave a high figure for the number of people having to sell houses to pay for care home fees. A little common sense logic showed that the figure had to be wrong. The numbers had come from an online survey that used a panel of respondents. The demographics of the respondents did not match up with likely numbers of parents in care homes and the issue of these online respondents being "happy clickers"? is discussed. Also discussed is the issue of double-counting.

North Sea helicopter safety: Differences between countries in the number of helicopter accidents suggest that action could be taken. However, it is explained that because there are so few accidents, it is impossible to say whether differences between countries are due to random fluctuation or not. It is suggested that accident rates per passenger mile are bound to be higher for North Sea helicopters than other land-based modes of transport and that this is an accepted part of being a worker in the North Sea.

Telephone waiting times: The amount of time people spend waiting on Government telephone lines was claimed to be more than it should have been. Even after correcting this, the number sounds very large. However, after adjusting for the number of people and the length of time, the number sounds less extreme.

<font size="+1">Episode 3 (13/09/2013)

URL: http://www.bbc.co.uk/programmes/b039rwd0

<font size="+1">Episode 4 (20/09/2013)

URL: http://www.bbc.co.uk/programmes/b03b2zbk

Evidence for the benefits of free school meals: Free school meals for young children are being introduced to schools based on pilot studies. Evidence from the pilots was not fully understood and the planning of pilots meant they were not ideal for drawing conclusions about the whole country and to "the average"? school. The Hawthorne effect is discussed as this might have been an issue in the pilots.

<font size="+1">Episode 5 (27/09/2013)

URL: http://www.bbc.co.uk/programmes/b03bg4w2

Hospital Standardised Mortality Ratios (HSMRs): The use of HSMRs to compare countries are discussed. The methods for comparing countries is not transparent and differences in the way healthcare is implemented in different countries means that in-hospital mortality ratios are bound to be different, irrespective of the quality of care. See also Radio 4 Series 29 Episode 1 (30/08/2013).

Hawthorne effect: The origins of the idea of the Hawthorne effect are discussed - how collecting data from people can change their responses. The original study is examined and evidence for the Hawthorne effect is questioned. See also Radio 4 Series 11 Episode 1 (23/10/2006) and World Service Episode 92 (12/10/2013).

<font size="+1">Episode 6 (04/10/2013)

URL: http://www.bbc.co.uk/programmes/b03bsb9y

Series 30 (Dec 2013 - Jan 2014)
<font size="+1">Episode 1 (20/12/2013)

URL: http://www.bbc.co.uk/programmes/b03lsdgj

<font size="+1">Episode 2 (27/12/2013)

URL: http://www.bbc.co.uk/programmes/b03m81b7

<font size="+1">Episode 3 (03/01/2014)

URL: http://www.bbc.co.uk/programmes/b03mjcbk

Europe's most improved railway: Posters at railway stations claimed that an international survey has ranked British railways as the most improved in Europe. While it emerges that the claim is true, it is also revealed that this is partially because they had ample room to improve, moving from 13th place to 10th place in the overall rankings where as other countries, such as Germany had less scope for improvement.

Clusters of deaths of cyclists: The deaths of six cyclists in London in a two week period prompted an investigation into how unusual this clustering was and whether it could point to a worsening of road safety for cyclists. Initial calculations suggest it is a very extreme event but once the number of possible two week periods in which this could occur is taken into account, the chances come down to a once in forty years probability - still unlikely but not so stunning. The assumption of constant road safety for cyclists over time to work out the calculations is also discussed.

<font size="+1">Episode 4 (10/01/2014)

URL: http://www.bbc.co.uk/programmes/b03nt8bf

<font size="+1">Episode 5 (17/01/2014)

URL: http://www.bbc.co.uk/programmes/b03pmk7p

<font size="+1">Episode 6 (24/01/2014)

URL: http://www.bbc.co.uk/programmes/b03qfzgx

Alcohol and mouth cancer: A leaflet suggests that the chances of getting mouth cancer are tripled for people drinking two glasses of wine per day. The distinction between absolute and relative risk is discussed. Also discussed is the problems of meta-analysis when studies from several countries with different alcohol consumption habits are being combined. The issue of smoking as a confounder is also discussed, suggesting that it is the interaction between alcohol and smoking which is most important. See also World Service Episode 107 (25/01/2014).

<font size="+1">Episode 7 (31/01/2014)

URL: http://www.bbc.co.uk/programmes/b03sb3x8

Series 31 (May - Jun 2014)
<font size="+1">Episode 1 (02/05/2014)

URL: http://www.bbc.co.uk/programmes/b041yf8s

Using Bayes to find an aeroplane: a slightly abbreviated version of the World Service broadcast on 22/03/2014. See also World Service Episode 115 (22/03/2014).

<font size="+1">Episode 2 (09/05/2014)

URL: http://www.bbc.co.uk/programmes/b042lp94

<font size="+1">Episode 3 (16/05/2014)

URL: http://www.bbc.co.uk/programmes/b0436h18

Probability with Willy Wonka's Golden Tickets: An exploration of probability using the example of Charlie Bucket finding one of Willy Wonka's Golden Tickets from the Roald Dahl book "Charlie and the Chocolate Factory". See also World Service Episode 131 (11/07/2014).

French people in London: A different, abbreviated version of the World Service broadcast. See also World Service Episode 116 (29/03/2014).

<font size="+1">Episode 4 (23/05/2014)

URL: http://www.bbc.co.uk/programmes/b04416sl

Probability and risk: A discussion of probability and how people find it hard to quantify risk and uncertainty. See also World Service Episode 124 (23/05/2014).

Spurious correlations: A discussion (including the use of sound as an aid) of spurious correlations and the issue of correlation not implying causation. The website http://tylervigen.com/ is referenced. See also World Service Episode 126 (06/06/2014).

<font size="+1">Episode 5 (30/05/2014)

URL: http://www.bbc.co.uk/programmes/b044jh75

Racism statistics: A newspaper headline suggested that racism was on the increase. However, a closer examination of the figures suggested that the rise could be spurious as the overall trend over a number of years was down. Includes discussion of moving averages and natural variability.

<font size="+1">Episode 6 (06/06/2014)

URL: http://www.bbc.co.uk/programmes/b045c660

Tossing a coin to help make decisions: A professor of Economics offered to help people make difficult decisions by tossing a coin for them. By doing this he gained an insight into how people make decisions and whether making life changes rather than sticking with the status quo enhanced people's happiness. Essentially he was creating a randomised trial in a situation where it would be impossible to plan one. See also Radio 4 Series 31 Episode 7 (13/06/2014) and World Service Episode 127 (13/06/2014).

Probabilities in office sweepstakes: The issue of choosing the optimum point to join an office sweepstake is discussed using the 2014 World Cup as an example. See also World Service Episode 126 (06/06/2014).

<font size="+1">Episode 7 (13/06/2014)

URL: http://www.bbc.co.uk/programmes/b0460zn1

Charitable giving: A survey suggested a link between religious belief and levels of charitable giving. The nature of this link is investigated and issues of sample size, causation are discussed along with the framing effect caused by the order in which questions are asked.

Big Data: Further discussion of issues surrounding Big Data, including Google Flu Trends, the 1936 US Presidential Election and finding pot holes in the roads of Boston. See also World Service Episode 63 (23/03/2013), Radio 4 Series 29 Episode 1 (30/08/2013) and World Service Episode 145 (17/10/2014).

Tossing a coin to help make decisions: Correspondence from listeners regarding the broadcast on 06/06/2014. See also Radio 4 Series 31 Episode 6 (06/06/2014) and World Service Episode 127 (13/06/2014).

Birthday paradox: A well-known apparent paradox in Statistics is that you only need to have 23 people in a room before the chances of two of them having the same birthday becomes more than 0.5. This issue is discussed using the birth dates of players in the 2014 World Cup squads. See also World Service Episode 127 (13/06/2014).

Series 32 (Aug - Sep 2014)
<font size="+1">Episode 1 (15/08/2014)

URL: http://www.bbc.co.uk/programmes/b04d4v85

Machine learning: A discussion of machine learning including methods used to win kaggle.com competitions.

<font size="+1">Episode 2 (22/08/2014)

URL: http://www.bbc.co.uk/programmes/b04dqyh4

<font size="+1">Episode 3 (29/08/2014)

URL: http://www.bbc.co.uk/programmes/b04fcsyt

Cyclists and guide dogs: A news report stated that 25% of guide dogs in London had been involved in collisions with cyclists. This claim was investigated and it emerged that the figure was from a survey with a self-selecting sample.

<font size="+1">Episode 4 (05/09/2014)

URL: http://www.bbc.co.uk/programmes/b04g1drf

Measuring anti-Semitism: An abbreviated version of the World Service version broadcast on 15/08/2012 is broadcast on Radio 4, omitting information on the methodology used for measuring and the dichotomisation issue. See also World Service Episode 136 (15/08/2014).

Measuring vocabulary: Comparing the size of vocabularies of rappers and Shakespeare using token analysis and stylometrics. See also World Service Episode 140 (12/09/2014).

<font size="+1">Episode 5 (12/09/2014)

URL: http://www.bbc.co.uk/programmes/b04gcfml

Political opinion polls using internet panels: In the run up to the Scottish independence referendum in 2014, one opinion poll put the "Yes" campaign marginally ahead. This caused great excitement but also raised the issue of how much one can trust polls based on internet panels. A discussion of internet panel surveys versus traditional surveys takes place. Also discussed are ways polling companies make adjustments when they know that certain responses may be perceived as unpopular and the potential infiltration of panels by certain groups.

Over half of Scotland living on benefits: A claim by Nigel Farage, leader of the UK Independence Party, that over half the people in Scotland are receiving state benefits is investigated, the implication being that these people are poor and Scotland is being subsidised by the rest of the UK. The claim is only true if one includes pensions, child benefits, tax credits, etc. and would apply to all other parts of the UK as well.

<font size="+1">Episode 6 (19/09/2014)

URL: http://www.bbc.co.uk/programmes/b04gyqtg

<font size="+1">Episode 7 (26/09/2014)

URL: http://www.bbc.co.uk/programmes/b04hywws

Proportion of 16 and 17 year olds voting for Scottish independence: After the Scottish independence referendum it was claimed that 71% of 16 and 17 year olds had voted "yes". However, it was revealed that although this came from a survey of 2000 people, only 14 of these had been 16 or 17 and 10 of these had voted "yes". The lack of precision that this gives was discussed.

Deaths from air pollution: Methods for quantifying early deaths from air pollution are discussed, including numbers of people dying early, number of years lost, etc. The inappropriate reporting of the figures and inappropriate comparisons with deaths in London smogs are also commented upon.

Series 33 (Jan - Feb 2015)
<font size="+1">Episode 1 (02/01/2015)

URL: http://www.bbc.co.uk/programmes/b04vqx64

<font size="+1">Episode 2 (09/01/2015)

URL: http://www.bbc.co.uk/programmes/b04xf1d5

National positions in league tables: An alternative version of the World Service version broadcast on 15/11/2014: An apparent increase in the number of league tables of countries is investigated. Some are identified as having good underlying methodologies whereas others have poor methodologies. See also Radio 4 Series 13 Episode 6 (03/12/2007), Radio 4 Series 13 Episode 7 (10/12/2007) and World Service Episode 149 (15/11/2014).

<font size="+1">Episode 3 (16/01/2015)

URL: http://www.bbc.co.uk/programmes/b04y53fk

Measuring reductions in wildlife: A shorter version of the World Service version broadcast on 10/10/2014: WWF and ZSL produced a report which claimed that the number of animals in the wild had reduced by 50% between 1970 and 2010. Two issues with this were identified. Firstly, the 50% figure was the result of averaging percentages so in reality the number of animals had not gone down by 50%. The fictitious example is given of 1970 figures for frogs being 10 and birds 100. If 2010 figures were 2 for frogs (80% reduction) and 80 for birds (20% reduction) then an average of 50% was being claimed when total numbers had reduced from 110 to 82. Secondly, the report had taken surveys of animal numbers from all over the world and weighted them to take account of the amount of the world covered by these surveys. However, concern was expressed that surveys may be more targeted towards threatened populations and thus the weighting applied might not be appropriate. See also World Service Episode 144 (10/10/2014).

<font size="+1">Episode 4 (23/01/2015)

URL: http://www.bbc.co.uk/programmes/b04yk0jf

<font size="+1">Episode 5 (30/01/2015)

URL: http://www.bbc.co.uk/programmes/b050c4sj

Predicted dates of birth: Two pregnant women have been given the same predicted date for giving birth but what are the chances of them both giving birth on the same day? For an individual mother, the probability of giving birth on the due date is 4.4% so the chances of them both giving birth on the due date is 4.4%*4.4% = 1 in 300. Other figures that are quoted is that the chances of birth occurring four days after the due date is slightly more than 4.4%, there is a 94% chance of birth occurring within two weeks of the due date and the chances of the two women giving birth on the same day (but not necessarily the due date) is 1 in 30. See also World Service Episode 160 (31/01/2015), World Service Episode 277 (28/04/2017).

<font size="+1">Episode 6 (06/02/2015)

URL: http://www.bbc.co.uk/programmes/b0512ln0

Strenuous jogging is bad for you: Reports claim that a study has shown that strenuous jogging is worse for health than no exercise at all. However upon investigation it turned out that the study contained just two strenuous joggers who had died. The dangers of extrapolating from small numbers is discussed. See also World Service Episode 161 (07/02/2015), World Service Episode 167 (21/03/2015), Radio 4 Series 34 Episode 1 (01/05/2015).

Series 34 (May - Jun 2015)
<font size="+1">Episode 1 (01/05/2015)

URL: http://www.bbc.co.uk/programmes/b05sfd0r

Strenuous jogging is bad for you: An update on this story where the researchers concede that they did not have enough evidence to draw the conclusions they did about strenuous jogging. See also Radio 4 Series 33 Episode 6 (06/02/2015), World Service Episode 161 (07/02/2015), World Service Episode 167 (21/03/2015).

<font size="+1">Episode 2 (08/05/2015)

URL: http://www.bbc.co.uk/programmes/b05syb68

<font size="+1">Episode 3 (15/05/2015)

URL: http://www.bbc.co.uk/programmes/b05tpz78

<font size="+1">Episode 4 (22/05/2015)

URL: http://www.bbc.co.uk/programmes/b05vhkc0

<font size="+1">Episode 5 (29/05/2015)

URL: http://www.bbc.co.uk/programmes/b05w8dnj

<font size="+1">Episode 6 (05/06/2015)

URL: http://www.bbc.co.uk/programmes/b05wz90d

Chocolate consumption and weight loss: A randomised control trial (RCT) was conducted using 3 groups - control, low carbohydrate diet and low carbohydrate diet with daily chocolate. Results showed weight loss in the two low carbohydrate groups with greater loss in the group where chocolate was also eaten. A paper was published in an online journal and a press release was issued, leading to stories in newspapers. However, it was a deliberately gad study with only 5 people per group and many potential outcomes measured. The paper did not explain these shortcomings and was submitted to a journal which it was believed would not carry out appropriate peer review. This highlights the issues of multiple testing and sample size as well as the importance of good peer review and standards of some journals. Also discussed are ethical issues connected with allowing the story to reach the popular press and whether journalists should be expected to assess the worth of scientific articles or trust the scientific literature. See also World Service Episode 179 (12/06/2015).

Probability in GCSE Maths: A GCSE Maths paper in 2015 contained a question combining probability and algebra. The question is explained and answers given. See also World Service Episode 178 (05/06/2015).

<font size="+1">Episode 7 (12/06/2015)

URL: http://www.bbc.co.uk/programmes/b05xqbml

Series 35 (Aug - Sep 2015)
<font size="+1">Episode 1 (14/08/2015)

URL: http://www.bbc.co.uk/programmes/b0659q1f

<font size="+1">Episode 2 (21/08/2015)

URL: http://www.bbc.co.uk/programmes/b065xk1v

CSIs at even numbered houses: Leicestershire Constabulary realised that sending a Crime Scene Investigator (CSI) to every burglary was not a very efficient use of resources. At the same time they were concerned that not sending them might have an impact on public confidence. They thus conducted a study to examine the impact of only sending CSIs to crime scenes where it was clear that there was evidence to collect. They did this by getting investigating officers to make a judgment about whether or not there was clearly evidence to be collected (or the victim was vulnerable) and if not, only sending a CSI to even numbered houses. In the discussion they are praised for realising that it is a good idea to do a study like this but it is pointed out that the trial was not blinded as the investigating officer's decision about whether it was appropriate for the burglary to enter the trial was made with knowledge of what "treatment" would then be applied. In the end the study was curtailed due to bad publicity but at that point, out of 96 houses, only 26 were odd numbers which might be down to the lack of blinding.

<font size="+1">Episode 3 (28/08/2015)

URL: http://www.bbc.co.uk/programmes/b06707kd

<font size="+1">Episode 4 (04/09/2015)

URL: http://www.bbc.co.uk/programmes/b06810qc

Queuing in reverse: It is explained that a queuing system where the last person in the queue was always served first could be the most efficient method from an overall time point of view. However it is also explained that it would be impractical in most situations. See also World Service Episode 191 (04/09/2015).

<font size="+1">Episode 5 (11/09/2015)

URL: http://www.bbc.co.uk/programmes/b06950lm

<font size="+1">Episode 6 (18/09/2015)

URL: http://www.bbc.co.uk/programmes/b06b374x

<font size="+1">Episode 7 (25/09/2015)

URL: http://www.bbc.co.uk/programmes/b06c4k46

Dementia - incidence and prevalence: A report suggested that the one-third of British people born in 2015 would develop dementia. The practicalities of predicting health outcomes so far in the future are discussed as are the definitions of the "incidence" and "prevalence" of a disease.

Reproducibility project: A project to try and reproduce the results of 100 psychology experiments concluded that only 39% successfully replicated the outcomes and many effect sizes were half what the original paper showed. A discussion of selective publishing, selective analysis and publication bias and possible solutions occurs. See also World Service Episode 194 (25/09/2015).

Series 36 (Jan - Feb 2016)
<font size="+1">Episode 1 (08/01/2016)

N.B. This episode was introduced as the beginning of a new series "unless you count our New Year's Day programme" (which had been introduced as a "special edition"). See Numbers of the Year 2015.

URL: http://www.bbc.co.uk/programmes/b06wbghr

<font size="+1">Episode 2 (15/01/2016)

URL: http://www.bbc.co.uk/programmes/b06vn925

Deaths from stroke at weekends: Government ministers promoting changes to working practices in the NHS were using figures that suggested there was an increased risk of dying in hospital from a stroke at weekends compared with weekdays. The issues around this are discussed including the fact that people who go to hospital at the weekend are likely to be more seriously ill than those going during the week.

Political opinion polls and the 2015 UK General Election: The result of the 2015 UK General Election was a surprise because the opinion polls just before the vote predicted less support for the Conservative party than occurred in the actual vote. Possible reasons for this are discussed following an analysis of data from the subsequent British Social Attitudes survey. It was concluded that the sampling was not good enough with Labour voters being easier to contact than Conservative voters. See also special edition Numbers of the Year 2015.

<font size="+1">Episode 3 (22/01/2016)

URL: http://www.bbc.co.uk/programmes/b06wv9cg

<font size="+1">Episode 4 (29/01/2016)

URL: http://www.bbc.co.uk/programmes/b06yfypf

Dangers of alcohol consumption: Following the issuing of stricter advice on limiting alcohol consumption, the difficulties of conducting research into this topic are discussed. It is difficult to get a sample of non-drinkers to compare with drinkers because non-drinkers are often in this category as a result of reasons such as previous over-consumption or religion/abstention. Also discussed are the difficulties of collecting reliable data as people under report consumption. See also World Service Episode 220 (25/03/2016).

Life table history and discrimination: Some history of how life tables for actuarial use were created in the USA (including how it was important to obtain data from relevant comparison populations). Also how comparison of life tables for different sections of the population revealed discrimination in action.

<font size="+1">Episode 5 (05/02/2016)

URL: http://www.bbc.co.uk/programmes/b06zcg4v

E-cigarettes and quitting smoking: A study suggested that using e-cigarettes actually had a detrimental effect on trying to stop smoking. This was a meta-analysis and was criticised by many for combining studies that did not look at the same issue (some were looking at users of e-cigarettes who were trying to quit and some were looking at all users of e-cigarettes). See also World Service Episode 213 (05/02/2016).

Birthday probabilities: A listener asked about the probability of two of his three children having the same birthday as him. The probability is calculated which includes having to allow for either parent matching and any two of three children matching. The slightly uneven spread of births across the year is also discussed. See also World Service Episode 214 (12/02/2016).

<font size="+1">Episode 6 (12/02/2016)

URL: http://www.bbc.co.uk/programmes/b06zvbwz

Time spent taking selfies: Various media outlets reported a story that the average woman aged 16-25 spent over 5 hours a week taking selfies. The survey was carried out by a company selling make-up and only asked women who took selfies. 2000 women were surveyed of whom 447 were aged 16-25. They found the average number of selfies taken per day and average time taken per selfie before multiplying up to get a figure for a week. However, some responses for number of selfies per day were incompatible with the mean time per selfie (they must have taken far less time than average). Also, modal categories were far less than the mean figures, suggesting skewed distributions and/or influential outliers.

Kevin McConway's Statistical Dictionary: Kevin McConway explains the technical meaning of everyday words used in Statistics: significance, error and regression (the latter including its origins in "regression to the mean").

Series 37 (Apr - May 2016)
<font size="+1">Episode 1 (01/04/2016)

URL: http://www.bbc.co.uk/programmes/b074zy97

Collecting and analysing food consumption data: The World Service broadcast of 11/03/2016 is repeated for Radio 4. The difficulties of collecting data on food consumption, including the Food Frequency Questionnaire and trying to verify its findings with the 24 Food Recall. The questionable nature of some research into food is demonstrated using the multiple comparisons problem and examples of spurious correlation and correlations caused by associations with a third variable. See also World Service Episode 218 (11/03/2016).

<font size="+1">Episode 2 (08/04/2016)

URL: http://www.bbc.co.uk/programmes/b075thgl

History of the mean: A history of the mean including where it was first used and the idea of the "average man". See also World Service Episode 222 (08/04/2016).

<font size="+1">Episode 3 (15/04/2016)

URL: http://www.bbc.co.uk/programmes/b076prgl

Opinion poll of UK Muslims: A Channel 4 programme commissioned a poll to gather opinions of UK Muslims that was carried out in areas of high Muslim residency, presumably for reasons of cost. The programme presented the results as being representative of all UK Muslims when it would have been more appropriate to represent them as being opinions of UK Muslims who live in areas where the Muslim culture is strong. Also discussed are methods that can be used to poll minority groups.

<font size="+1">Episode 4 (22/04/2016)

URL: http://www.bbc.co.uk/programmes/b077jqps

Life expectancy of Popes: A slightly abbreviated version of the World Service broadcast of 15/04/2016 is repeared for Radio 4. Survival analysis of historical ages of Popes at death. See also World Service Episode 223 (15/04/2016).

<font size="+1">Episode 5 (29/04/2016)

URL: http://www.bbc.co.uk/programmes/b0788bcb

Measuring Migration: The level of migration to the UK from other EU countries is discussed. The International Passenger Survey and requests for National Insurance numbers give different estimates but it is emphasised that they are trying to measure different things. See also the story on the International Passenger Survey broadcast on Radio 4 on 27/01/2005 (Radio 4 Series 7 Episode 3) and special edition The Referendum by Numbers, episode 2: Immigration.

Simpson's Paradox: Simpson's Paradox is discussed with reference to the discovery that at University of California, Berkeley, admission acceptance rates for males and females were roughly equal in each department but overall more males were being offered places than females. The reason for this paradox was that women were more likely to be applying for courses that had higher rejection rates, with the reverse being the case for males. The paradox is also discussed with reference to auditions for a fictitious choir. See also Radio 4 Series 7 Episode 3 (27/01/2005), World Service Episode 225 (29/04/2016), Radio 4 Series 41 Episode 1 (25/08/2017).

<font size="+1">Episode 6 (06/05/2016)

URL: http://www.bbc.co.uk/programmes/b0790cgp

Series 38 (Jul - Sep 2016)
<font size="+1">Episode 1 (29/07/2016)

URL: http://www.bbc.co.uk/programmes/b07lhl0n

Algorithms and sorting efficiency: A discussion of algorithms and their efficiency including whether or not it is worthwhile sorting some things like emails. See also World Service Episode 238 (29/07/2016).

<font size="+1">Episode 2 (05/08/2016)

URL: http://www.bbc.co.uk/programmes/b07m7z12

Measuring income inequality: The measurement of income inequality is discussed including the Gini coefficient and comparing different percentiles of income distribution. See also See also Radio 4 Series 24 Episode 4 (23/12/2011).

<font size="+1">Episode 3 (12/08/2016)

URL: http://www.bbc.co.uk/programmes/b07mz0hj

<font size="+1">Episode 4 (19/08/2016)

URL: http://www.bbc.co.uk/programmes/b07ns0qy

<font size="+1">Episode 5 (26/08/2016)

URL: http://www.bbc.co.uk/programmes/b07pjkj5

<font size="+1">Episode 6 (02/09/2016)

URL: http://www.bbc.co.uk/programmes/b07qc93b

Synchronisation of menstrual cycles: Anecdotal and past research says that women who live in close proximity are likely to synchronise their menstrual cycles. However, is the synchronisation observed more than what would happen by random chance? See also World Service Episode 243 (02/09/2016).

Black box algorithms: Algorithms that are used in everyday life and affect the way we live are discussed, particularly those whose internal workings are hidden from public view. Examples are given and ways of "gaming" the algorithms discussed. See also World Service Episode 251 (28/10/2016).

Series 39 (Nov - Dec 2016)
<font size="+1">Episode 1 (04/11/2016)

URL: http://www.bbc.co.uk/programmes/b081tq55

Worldwide wildlife numbers: The difficulties of averaging averages and combining data from different sources and locations are discussed.

<font size="+1">Episode 2 (11/11/2016)

URL: http://www.bbc.co.uk/programmes/b081tjh4

<font size="+1">Episode 3 (18/11/2016)

URL: http://www.bbc.co.uk/programmes/b082klf9

<font size="+1">Episode 4 (25/11/2016)

URL: http://www.bbc.co.uk/programmes/b08343jx

Finding a friend in a crowd: Simulations indicate that both people moving around means they are likely to meet sooner than if one of them stays still. Alternatively it might be better to concentrate on focus points and visit those locations.

<font size="+1">Episode 5 (02/12/2016)

URL: http://www.bbc.co.uk/programmes/b083r9x5

Danny Dyer's royal connection: The TV programme "Who Do You Think You Are?" revealed that one of Eastenders actor Danny Dyer's 22 x Great Grandfathers is King Edward III. Rather than being surprising, calculations show that it is actually very likely (although being able to reliably track this is unusual).

<font size="+1">Episode 5 (02/12/2016)

URL: http://www.bbc.co.uk/programmes/b083r9x5

Primary school assessments: It was revealed that a third of children in the early years of primary school are bring assessed as below the "expected standard". It is revealed that this is not surprising as the standards are set at a level that one would not expect all children to attain at that age. Also discussed are the incentives that targets create for these teacher observation based assessments. There are 17 criteria and the number of children that score 34 overall (equivalent to scoring 2, the "expected" level, on all criteria) is very large compared with other scores, even close to 34. It is explained that it is good for a school to have large proportions of children obtaining the "expected standard" but because assessments in subsequent years also look at progress (in relation to these early assessments), it is advantageous for the school not to assess the children as achieving too highly.

<font size="+1">Episode 6 (09/12/2016)

URL: http://www.bbc.co.uk/programmes/b084dqpr

<font size="+1">Episode 7 (16/12/2016)

URL: http://www.bbc.co.uk/programmes/b08504ln

<font size="+1">Episode 8 (23/12/2016)

URL: http://www.bbc.co.uk/programmes/b085bfys

Series 40 (Apr - Jun 2017)
<font size="+1">Episode 1 (14/04/2017)

URL: http://www.bbc.co.uk/programmes/b08lk542

<font size="+1">Episode 2 (21/04/2017)

URL: http://www.bbc.co.uk/programmes/b08mdcj0

<font size="+1">Episode 3 (28/04/2017)

URL: http://www.bbc.co.uk/programmes/b08n4lkg

<font size="+1">Episode 4 (05/05/2017)

URL: http://www.bbc.co.uk/programmes/b08ns2lw

Recording crime: Crime rate trends differ across the UK and across years. Reasons for this are examined, including recording practices.

<font size="+1">Episode 5 (12/05/2017)

URL: http://www.bbc.co.uk/programmes/b08pgswp

Penalty shoot-out sequences: It has been observed that teams that go first in penalty shootouts where turns alternate are more likely to win (around 60% of the time). A trial is being carried out to a ABBA,ABBA,ABBA,… system so that the psychological advantage changes. This is still not entirely fair as B is always catching up with A's advantage and never has a greater advantage than A. The Thue-Morse sequence ABBA,BAAB,BAAB,ABBA,... is an even fairer system as the advantage alternates, but there are questions as to the confusion it might create. See also World Service Episode 286 (30/06/2017).

<font size="+1">Episode 6 (19/05/2017)

URL: http://www.bbc.co.uk/programmes/b08q7l88

<font size="+1">Episode 7 (26/05/2017)

URL: http://www.bbc.co.uk/programmes/b08r320w

Pre-release access to statistics for UK Government ministries: Following suspicious movements in the financial markets, a study was undertaken to see if they could be due to leaks from the pre-release access to statistics for UK Government ministries. A relationship was found whereas a similar study in Sweden, where there is no pre-release access, did not find any relationship. A number of bodies, including the Royal Statistical Society have been campaigning for this pre-release access system to end, not just because of potential issues in the financial markets but because of the unfairness it brings to politics. (N.B. Pre-release access was ended in June 2017.)

<font size="+1">Episode 8 (02/06/2017)

URL: http://www.bbc.co.uk/programmes/b08sl5xt

Series 41 (Aug - Sep 2017)
<font size="+1">Episode 1 (25/08/2017)

URL: http://www.bbc.co.uk/programmes/b091wf7v

Simpson's Paradox: A greater percentage of boys are getting A and A* grades at A-level than girls. However, boys tend to be doing subjects which give a larger proportion of A and A* grades than those subjects which tend to be preferred by girls. In part, the programme returns to the Radio 4 broadcast of 29/04/2016 and World Service broadcast of 02/05/2016. See also Radio 4 Series 7 Episode 3 (27/01/2005), Radio 4 Series 37 Episode 5 (29/04/2016), World Service Episode 225 (29/04/2016).

<font size="+1">Episode 2 (01/09/2017)

URL: http://www.bbc.co.uk/programmes/b092ggjk

<font size="+1">Episode 3 (08/09/2017)

URL: http://www.bbc.co.uk/programmes/b092rbc0

Funding Mathematics and Statistics research: A discussion of the difficulties of Mathematics and Statistics being appreciated and funded in a climate where short term impact is considered to be most important.

<font size="+1">Episode 4 (15/09/2017)

URL: http://www.bbc.co.uk/programmes/b093hf8v

<font size="+1">Episode 5 (22/09/2017)

URL: http://www.bbc.co.uk/programmes/b0953bdx

<font size="+1">Episode 6 (29/09/2017)

URL: http://www.bbc.co.uk/programmes/b095tn27

Measuring domestic violence: A discussion of how levels of domestic violence can be measured. The Crime Survey for England and Wales is used and the discussion includes the issues of data capping (using a maximum for number of incidences to avoid unusual cases skewing summary statistics) and collecting intimate questions.

Series 42 (Jan - Feb 2018)
<font size="+1">Episode 1 (12/01/2018)

URL: http://www.bbc.co.uk/programmes/b09l2351

Average graduate salaries: The "graduate premium" is discussed - the added amount that a graduate might be expected to learn, compared with a non-graduate. Comparisons of median wages are used although the hazards of using the mean are not discussed. The fact that there are more graduates in recent decades would lead standard economic theory to suggest that the graduate premium would be reduced. This is not the case and reasons of productivity are discussed. Another study used anonymised UK tax data and student records to examine the same issue. It suggested that some subgroups had greater graduate premiums than others and this was also true for different degree courses and universities. It is acknowledged that different degrees and universities had different entry requirements and thus the groups being compared are different in characteristics beyond just wages. Once this is allowed for, there are still differences between universities with some institutions having negative graduate premiums. In relation to this last point, the story concludes by reminding listeners that where graduates lived made a big difference to wages and that there are large differences in wage prospects for different degree courses. The unstated subtext is that amongst those institutions with a supposed negative graduate premium, there will be some that are based in lower-wage areas of the country whose local students may have wages less than the average for non-graduates in the whole country but greater than the average for non-graduates in their locality. Also, some of the institutions with supposed negative graduate premiums may have high proportions of graduates who subsequently work in sectors which have lower pay such as the arts. The issue is thus about the need to make comparisons between relevant groups rather than compare subgroups with the overall average. See also Radio 4 Series 11 Episode 6 (27/11/2006).

<font size="+1">Episode 2 (19/01/2018)

URL: http://www.bbc.co.uk/programmes/b09m198x

<font size="+1">Episode 3 (26/01/2018)

URL: http://www.bbc.co.uk/programmes/b09p4fc4

<font size="+1">Episode 4 (02/02/2018)

URL: http://www.bbc.co.uk/programmes/b09ppxc0

Accuracy of the groundhog: The World Service Special Edition of 18/02/2012 is repeated as a bonus podcast item, in addition to the usual broadcast. See http://moreorless.wikia.com/wiki/Browse_the_World_Service_episodes World Service Episode 6 (Special Edition) (18/02/2012)].

<font size="+1">Episode 5 (09/02/2018)

URL: https://www.bbc.co.uk/programmes/b09qhw6t

<font size="+1">Episode 6 (16/02/2018)

URL: https://www.bbc.co.uk/programmes/b09r82r3

Series 43 (Apr - Jun 2018)
<font size="+1">Episode 1 (27/04/2018)

URL: https://www.bbc.co.uk/programmes/b0b01w69

<font size="+1">Episode 2 (04/05/2018)

URL: https://www.bbc.co.uk/programmes/b0b0xll3

<font size="+1">Episode 3 (11/05/2018)

URL: https://www.bbc.co.uk/programmes/b0b1tmb2

Estimating the extent of modern slavery: The extent of modern slavery in the UK is difficult to estimate because of its illegal nature. A discussion of how modern slavery is defined takes place. A number of organisations have records of individuals who have been exploited in this way and this can be used to produce an estimate of the overall number of people in this situation. Methods related to capture-recapture methods commonly used in estimating wildlife populations can be used.

<font size="+1">Episode 4 (18/05/2018)

URL: https://www.bbc.co.uk/programmes/b0b2mrcf

<font size="+1">Episode 5 (25/05/2018)

URL: https://www.bbc.co.uk/programmes/b0b3fz4c

<font size="+1">Episode 6 (01/06/2018)

URL: https://www.bbc.co.uk/programmes/b0b493vt

<font size="+1">Episode 7 (08/06/2018)

URL: https://www.bbc.co.uk/programmes/b0b53dmy

Exam question probabilities: A history student is due to sit an exam which will contain 12 questions, each on a topic from a set of 20 topics. The student will have to choose 3 questions to answer and wants to know how few topics to revise in order to have a 90% chance of being able to answer 3 questions. The binomial probabilities concerning this scenario are calculated and discussed along with an approximation for the calculation.

Go back to More Or Less Wiki main page