Publications by Social data blog

Problems with R

20.09.2010

Struggling for two whole days to do some not-terribly-advanced repeated-measures analysis of variance in R. Probably my worst experience with R so far and really an example of where R could do with some centralisation of effort. I have explored six or seven ways to do more or less the same thing, but each approach was too lacking either in docum...

1188 sym

What is wrong with this graph?

28.09.2010

This is a graph from our surveys with the University of Munich in B&H about ten years ago.(Click on it to see a large version.) In previous presentations (but not in peer-reviewed journals) I have talked about the peak on the subscale of Paranoid Ideation for the B&H samples. But looking at it again I see that this is partly an illusion and it i...

1036 sym 2 img

Students in predominantly ethnic minority classes want segregated education very much. The others don’t.

07.10.2010

I just did this for what will hopefully be a book chapter on our Divided Education – Divided Citizens research project with NEPC. Explanation further below for anyone more interested in the actual topic 😉 About the graphic: I like raw-data plots like this, made possible by Hadley Wickhams’s amazing ggplot2 package for the stats package R....

3492 sym 2 img

Boxplots or raw data graphs?

14.10.2010

We recently had a dilemma for an OSI publication about the design for the graphs. There will be dozens of these graphs showing the mean score on a given variable for nearly 11000 parents from 10 countries. This example is for household wealth which has values ranging from 0 to 16. These are the three alternative designs we considered, all constru...

1553 sym 6 img

Which parents are satisfied with their child’s education? Those who know their class representative well. Especially in poor schools.

14.10.2010

Another result from our OSI / ESP survey of nearly 11000 parents in ten countries. Dots are individual parents.The y-axis is individual parents' overall satisfaction with their children's education. Red dots are parents who know their parent representatives well, blue dots are parents who do not; i.e. colour is mapped to the variable par.rep.1 wh...

1732 sym 2 img

Are class representatives twice as satisfied?

15.10.2010

So Agnes asked (via Facebook), looking at my previous post, http://stevepowell99.posterous.com/which-parents-are-satisfied-with-their-childs, if class representatives are more satisfied than other parents? This picture shows that they are: the parents with Rep=TRUE, i.e. the representatives, are also more satisfied, especially in schools where pa...

1151 sym 2 img

Looping through a set of graphics in odfWeave

03.11.2010

At proMENTE social research we often use the odfWeave and Sweave packages for the amazing statistics program R for automating the production of graphics and reports. odfWeave and Sweave are for the OpenOffice and lyx (www.lyx.org) word processors respectively. One problem with this approach arises when you have to produce and insert into your doc...

2149 sym R (191 sym/2 pcs)

Also in Bosnia & Herzegovina, satisfied parents are those who believe their school representatives are effective

11.11.2010

In the mixed effects regression for all 10 countries in the research I mentioned previously on this blog http://stevepowell99.posterous.com/which-parents-are-satisfied-with-their-childs, while school mean of “parents believing their representatives are effective” was significantly positively related to their overall satisfaction with educatio...

2538 sym 2 tbl

Keeping R libraries in sync between different computers using Dropbox

11.12.2010

We have a few computers including laptops in our network which all use R (r-project.org) for statistics. We use Dropbox to keep all our files in sync and we are all on ubuntu. The problem was that we wanted to keep our R installations in sync so we don’t have different libraries and settings everywhere. All you need to do is copy the Renviro...

1705 sym

My first Reproducible Research Compendium

14.12.2010

I have just completed my first Reproducible Research Compendium “Analysis of the combined survey datasets from the American Red Cross Tsunami Recovery Program Psycho-Social Project (adult community respondents)”. It is basically all the reports and data from all the work I did on evaluation psychosocial projects for the American Red Cross, bu...

4624 sym