We often receive feedback from participants that they are unaware of what ALSPAC does with the data it collects, what results have come from the study, or why some of the questions are so 'weird'! ResearchFest 2012 is ALSPAC's first event which tries to answer some of these questions.

This application is to use anonymised datasets to demonstrate to participants in an engaging and visual way what data can look like. Specifically we aim to explain:

1. correlation

2. confounding

3. why we have asked certain questions/ collect certain information:

a) alcohol use* - contains percentages of girls and boys with certain drinking habits at age 10-15. We would like to contrast use in girls and boys as although boys claim to drink more at age 10, girls are drinking more at age 13 and 15

b) birthweight - this is the file which illustrates confounding. We plan to create a line/scatter plot of "coffee" (cups of coffee) with "both" (average birthweights of the children of both smokers and non-smokers combined) to see an inverse association between coffee drinking during pregnancy and birthweight. If you "adjust" for smoking during pregnancy, i.e. stratify on smokers/nonsmokers this association between coffee drinking and birthweight largely goes away i.e. can do a line/scatter plot of "coffee" with "smoker" and "coffee" with "non-smoker".

c) reasonforsex - again, we thought it would be interesting to contrast the different reasons boys and girls had for having sex at age 15 so have included relative proportions of the different answers for boys and girls separately

d) riskbehaviour15_16* - table of various risk behaviours and proportions of individuals (total, boys and girls) partaking in them. We will either contrast boys/girls; or perhaps make the participants guess the % of individuals who had partaken in such risk behaviours at age 15-16 (some of the % ages are suprising - e.g. 42.1% of individuals reported criminal offending or antisocial behaviour. Other data on drug use and sexual activity is also in this file.)

e) tannergrp - this is the "lie-detector" file which shows how boys generally lied about their stage of pubertal development when asked at a young age. We would like to see if we could develop the programme so that participants could plot on a graph the age at which they think boys go through each of the stages. This can then be contrasted with the different trends I have compiled for:

- boys who claimed they were at tanner stage 4 when they were age 8 (which shows that these boys seem to go backwards in their stage of development when asked at later time points)

- boys who claimed they were at tanner stage 1 when they were age 8 (this is the anticipated stage at this age)

- average stage reported by boys in ALSPAC at all ages

- population-average tanner stage at each age

(basically what these trends show is that boys in ALSPAC generally exaggerated their stage of penis development compared with what is expected at the different ages).

[*both these data sets are taken directly from a publication: Patterns of alcohol use and multiple risk behaviour by gender during early and late adolescence: the ALSPAC cohort - MacArthur et al, 2012]

f) "talking fridges" The data collected to provide the PLIKS score remain memorable to many YPs. We would like to highlight that each question individually might seem odd, but together they can show something interesting that correlates with other things e.g. IQ and mothers smoking during pregnancy.

LH, LP, RR and MT are all direct users of ALSPAC data and willing to collate the information so no data buddy will be required.