the Chart Builder only computes means over quantitative variables.next, FORMATS adds percent signs to the recoded values.since we want to see percentages instead of proportions, we'll RECODE our 0-1 variables into 0-100.However, a couple of adjustments are still desirable: Our data now have the structure required for running our chart. If everything went right, the 2 checktables will show the exact same information. do repeat #target = work_1 to work_8 / #source = leis_1 to leis_8. *Shift values from leisure into work variables. value labels Purpose 0 'Work' 1 'Leisure'. If you're not sure about some command, you can inspect its result in data view if you run EXECUTE right after it. Let's now run these steps on transportation.sav with the syntax below.
This method saves more effort insofar as you stack more variables: placing the IF command in step in a DO REPEAT loop does the trick. The screenshots below illustrate how to mimic VARSTOCASES without needing its syntax. We'll replace it with a trick that creates nicer results and requires less syntax too. Now, we could use VARSTOCASES for our data but we find this rather tedious for 8 variables. The usual way to do so is with VARSTOCASES as illustrated below. For creating our chart, we need to stack these 2 sets vertically. Note that our data consist of 2 sets (work and leisure) of 8 dichotomous variables (transportation options). The data thus obtained are in transportation.sav, partly shown below. Respondents could select one or more means of transportation for both work and leisure related travelling. “which means of transportation do you use on a regular basis?” Example DataĪ sample of N = 259 respondents were asked As this requires restructuring our data, we'll first do so with a seriously cool trick. This tutorial shows how to create the clustered bar chart shown below in SPSS. Clustered Bar Chart over Multiple Variables By Ruben Geert van den Berg under Charts in SPSS
0 kommentar(er)
