Exercise 1.2. Data exploration
In order to get to know the data, we start with some data exploration.
- Take a look at the means and standard deviations for the three anti-immigration items. Calculate these statistics for the pooled dataset, i.e. without differentiating between countries or time points. Make sure that you have the correct dataset open. Also define DWEIGHT as a weighting variable.
Which item has the lowest mean? How do you explain that precisely this item has the lowest mean?
*Open the correct dataset. *Please do not forget to change ‘C:\’ to the path where you stored the ESS datasets.GET FILE='C:\ESS123_immig.sav'.
*Define ‘dweight’ as the weighting variable.WEIGHTBY dweight.
*Calculate means and standard deviations.MEANS IMSMETN IMDFETN IMPCNTR.
Table 1.2. Mean values of the three immigration items Variable Mean N Std. Deviation Allow many/few immigrants of same race/ethnic group as majority 2.23 84310 0.819 Allow many/few immigrants of different race/ethnic group from majority 2.50 84310 0.843 Allow many/few immigrants from poorer countries outside Europe 2.52 84310 0.852
Weighted by design weight.
IMSMETN has a lower mean score (2.23) than the other two items. The lower score reflects less opposition to immigration. Apparently, Europeans are less resistant to immigrants with the same ethnicity than to immigrants with a different ethnicity or from poor countries outside Europe (these two latter groups tend to largely overlap). It seems very plausible that attitudes towards immigrants who share some characteristics with the majority population are least negative.
- Graphs can be a very useful tool for data exploration. For each of the immigration items, draw a grouped bar chart representing the country means at the three time points. Your bar charts should represent the item averages (on the Y-axis) for all countries and time points (on the X-axis). The bar chart should have the following structure:
Figure 1.1: Example of a grouped bar chart
The bar charts can be obtained using SPSS syntax or SPSS ‘Chart Builder’.
*Draw a grouped bar chart representing the country-means at the three time points.GRAPH/BAR(GROUPED)=MEAN(imsmetn) BY cntry BY essround.GRAPH/BAR(GROUPED)=MEAN(imdfetn) BY cntry BY essround.GRAPH/BAR(GROUPED)=MEAN(impcntr) BY cntry BY essround.
* Chart created using Chart Builder.GGRAPH/GRAPHDATASET NAME="graphdataset" VARIABLES=cntry MEAN(imsmetn)[name="MEAN_imsmetn"] essround[LEVEL=NOMINAL] MISSING=LISTWISE REPORTMISSING=NO/GRAPHSPEC SOURCE=INLINE.BEGIN GPLSOURCE: s=userSource(id("graphdataset"))DATA: cntry=col(source(s), name("cntry"), unit.category())DATA: MEAN_imsmetn=col(source(s), name("MEAN_imsmetn"))DATA: essround=col(source(s), name("essround"), unit.category())COORD: rect(dim(1,2), cluster(3,0))GUIDE: axis(dim(3), label("Country"))GUIDE: axis(dim(2), label("Mean Allow many/few immigrants of same race/ethnic group as majority"))GUIDE: legend(aesthetic(aesthetic.color.interior), label("ESS round"))SCALE: cat(dim(3))SCALE: linear(dim(2), include(0))SCALE: cat(aesthetic(aesthetic.color.interior))SCALE: cat(dim(1))ELEMENT: interval(position(essround*MEAN_imsmetn*cntry), color.interior(essround), shape.interior(shape.square))END GPL.
Click ‘Graphs’ in the main horizontal tool bar, and then ‘Chart Builder’.
The Chart Builder wizard pops up. The first thing that needs to be done is to change the measurement level of the items from nominal to scale. This is necessary because SPSS does not allow you to calculate the mean of nominal variables. And it is precisely the item ‘means’ that we want to display on the Y-axis. Change the measurement level by right-clicking the variable name (in the upper left corner of the chart builder window) and then selecting ‘scale’ instead of ‘nominal’.
Next, we define the graph type. In the lower pane of the Chart Builder, select ‘bar’. Drag the second icon (with the green and blue bars next to each other) into the upper pane of the chart builder. Now you can select the variables to be displayed by dragging the variable names (in the upper left corner) into the graph. Drag IMSMETN to the box next to the Y-axis, and CNTRY to the X-axis. The variable ESSROUND should be dragged into the remaining box in the upper right corner (‘cluster on’ - see figure below). The measurement level of the variable ESSROUND must be set to ‘nominal’.
We still need one more option. Click ‘Element Properties’ in the upper part of the element properties window, select ‘X-axis (Bar 1)’. In the lower part of the screen, select ‘Show only categories present in the data’. Without this option, countries that were excluded from the dataset (but that still have their value labels present) would also be shown in the graph. Click ‘Apply’ and close the ‘Element properties’ window. To get the actual graph, click OK in the ‘Chart Builder’ window.
Repeat these steps for the other two items (IMDFETN, IMPCNTR).
- In which countries do we witness the strongest resistance to immigration? Which countries have the most immigration-friendly climate?
- Is there any evidence that attitudes toward immigration are changing over time? Do all countries under study experience similar evolution?
- Do the three items display a similar picture?
- Portugal and Hungary have the highest mean scores, indicating strong opposition to further immigration to the country. Swedes have by far the most positive attitudes to immigration.
- At first sight, attitudes toward immigration seem to be changing in some countries. The over-time evolutions are very different from one country to another, however. In some countries, such as Hungary, Portugal and the Netherlands, there seems to be a rather clear increase in anti-immigration attitudes. In others, such as Poland, we see the opposite trend.
- The items referring to immigrants of a different race/ethnic group (IMDFETN) or immigrants from poorer countries outside Europe (IMPCNTR) are very similar. This is not surprising, as there is a big overlap between these two immigrant groups. The item on immigrants from the same race/ethnic group (IMSMETN) leads to somewhat different conclusions, especially with respect to differences between countries. While Hungary clearly has the highest mean score on IMDFETN and IMPCNTR, Portugal scores highest on IMSMETN. The evolutions within countries, however, are largely similar for the three items.