STATISTICA Blogs

Do you ever wake up and say “I want to find some interesting data today”? 

Rather than acknowledge the fact that I might be the only geek here, I’ll just assume the answer is yes.

I knew the monthly jobs report was due out today, but I’ve already written about how “good” job report news isn’t always as positive as it seems on the surface.  

So I looked elsewhere and ended up at the World Bank. 

I drilled down to the Gender Statistics database and selected all the countries that were available in the list. Two of the interesting items were “Proportion of seats held by women in national parliaments” and “Wage equality between women and men for similar work (ratio),” which seemed they’d be intriguing to look at both individually and plotted together. 

Quick, before scrolling down, make an assumption about the relationship between wage equality and percentage of seats in national parliaments. Surely societies that elect many women to office will have mechanisms to ensure wage equality between the genders, right?

Were you wrong too? 

Rather than acknowledge the fact that I might be the only person who makes mistakes, I’ll just assume the answer is yes.

I created a Scatterplot with Histograms (Graphs menu -> 2D graphs -> Scatterplots w/ Histograms) so that we could look at the scatterplot of how the two variables relate to each other, but to also look at how the variables are distributed separately. 

Looking at the scatterplot, the points don’t seem to indicate a clear relationship between seats in parliament and wage equality. In fact, the linear fit line is sloping downward, which suggests that as the proportion of seats held by women increases, wage equality decreases. 

Looking above the scatterplot at the histogram for the proportion of seats in national parliaments, the greatest number of observations is between 10% and 15% of seats being held by women. Two observations were between 45% and 50% (Rwanda and Sweden). Eight countries reported that 0% of the seats were held by women.

The wage equality ratio histogram to the right of the scatterplot indicates that the highest number of observations was found between 60% and 65%. It was unclear from the World Bank’s description of the data set how this ratio was calculated, so I do not know what adjustments or assumptions were made. Eleven countries reported ratios above 80%, and there were also a few observations below 50%. 

After creating the graph, I wanted to see what the total counts for each variable were. I opened the Graph Options dialog box by double clicking on the graph, and I selected the Histogram node, clicked on the Statistics button, and checked the Total Count check box. This added the N for each variable to the title. The N for the proportion of parliament seats held by women was 186, and the N for wage equality was only 117. Therefore, if you want to make any serious conclusions taken from these data, you would first need to investigate why only 117 countries reported the wage equality data. Perhaps if these missing data were reported or estimated, the overall relationship between these two variables would appear different.  

 Have a great weekend, and remember to always look at and create graphs responsibly. 

 Image Credit: "Lynne Featherstone with Elay Ershad" posted by ukhomeoffice