Hello
I have two separate independent variables. Each variable was analysed at different levels. I give you an example: 1st Independent variable: the use of Facebook and LinkedIn
Levels: using Facebook and LinkedIn - being an active member - creating a group for interaction Each of these features had three choices: using facebook, using linkedIn, and using both. I measured the frequencies of these choices (how many said using Facebook, using LinkedIn, and using both), and I calculated the mean for each choice. Since the choices were the same in each level, it would be practical to compare the means of these choices (e.g., Comparing the means of using Facebook in the first level, in the second level, in the third level). If there significant differences between the means, the hypothesis is alternative. If equal, the hypothesis is null. The sample size: I had different groups (4 samples), and it would be impractical to compare the four samples because the size is not the same. Therfore I decided to combine these four groups, and I obtained a sample of 24 participants. Any suggestions, commnets, concerns ... you are welcome
You could use the Cochran Q test.
With SAS:
http://support.sas.com/documentation/cdl/en/procstat/63104/HTML/default/viewer.htm#procstat_freq_sect034.htm
In R, see attachment. It is for the data from Siegel & Castellan, Nonparametric Statistics for the Behavioral Sciences. 1988, 2nd Ed., p. 173.
For post hoc tests in R see, Salvatore S. Mangiafico :
http://rcompanion.org/handbook/H_07.html
Please check these:
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=2ahUKEwjNisiGh-TeAhWiM-wKHaIvCHIQFjADegQICRAC&url=https%3A%2F%2Fresources.vam.wfp.org%2Fsites%2Fdefault%2Ffiles%2F12%2520-%2520Hypothesis%2520Testing%2520with%2520Categorical%2520Variables%2520-%25202011.pptx&usg=AOvVaw2bwN8uQxwM9syqnnLmGYsv
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwjNisiGh-TeAhWiM-wKHaIvCHIQFjAAegQIABAC&url=http%3A%2F%2Fwww.mit.edu%2F~6.s085%2Fnotes%2Flecture6.pdf&usg=AOvVaw0vK_dFghhhkigI-JAKlzUS
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&uact=8&ved=2ahUKEwjNisiGh-TeAhWiM-wKHaIvCHIQFjAEegQIBhAC&url=http%3A%2F%2Fwww.pitt.edu%2F~nancyp%2Fstat-0200-s10%2Fslides%2FbpX1textfieldslecture31-catquanmoreanova.pdf&usg=AOvVaw0O3cGnllfw-eGAHRW-OIUi
My R script was wrong! Sorry.
In attachment the correct script. Now it gives the same results as in Siegel & Castellan p. 173
There are 3 columns (1 for Facebook, 1 for LinkedIn and 1 for Both), but you can expend these. (
Codes: 1 = yes, 0 = no
Results:
Column Sums:
FaceBook LinkedIn Both
13 13 3
Cochran test:
Q = 16.6666666666667 Degrees of Freedom = 2
The p-value is 0.000240369476419477
Jos, Noha, Matt Thank you very much. I will consider what you suggested, and see wich one fits in with the entire reserach process.
Bless you all!
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA