i am conducting a research study on effect of advertisement (six different types) on goal pursuit approach of individuals (2 goal pursuit approaches). type of advertisement in the research is dependent variable and goal pursuit approach is dependent variable. Whether 6x2 factorial design will be the right experimental method or Quasi experimental design with six different experiments ? Guidance is Required from the experts in the field.
Hello Shaukat,
The answer depends on several things. If you intend to assign cases at random to one of the six advertisement types, then, yes--it is an independent variable that may be handled as a between-subjects factor in an anova design. If cases are exposed to all six ad types (or are asked to self-report on exposure to the six ad types), then you'd have to treat that as a repeated measures (within-subjects) factor. If the values are linearly dependent (e.g., what percent of time is your exposure to ad type #3?), then you'd have to omit one of the six levels from the analysis.
If the DV is a choice, that is, persons opt for goal A or goal B, then that is a dichotomous variable, and anova or ordinary least squares regression would not be suitable. Logistic regression could be used.
If you have two DVs, and are quantifying the values for each, then possibly a manova could be run.
Good luck with your work.
David Morse Thank You for the suggestion, Yes i will be assigning subject randomly to one of the six advertisement format so thank you for suggesting between the subject. but for the two goal pursuit approaches can i go for with in subject effect in the same experiment, since i am measuring the effect of advertisement on two different goal pursuit approaches.
what i simply mean is that can i use between subject design for independent variable and with in subject design for dependent variable ?
Suggestion required please
Hello again Shaukat,
Yes, if the measures are otherwise comparable, this can be a repeated measures (within-subjects) factor. You'll get three principal tests: (a) a comparison of groups across the two scores; (b) a comparison of the two scores (goal pursuit methods) across the groups; and (c) a group by score interaction test. It is likely this last effect that will be of greatest interest.
Good luck with your study!
Hello. This will be a factorial research plan, and for calculations, accordingly, it is necessary to apply a one-way analysis of variance.
- Badges
- Science method
- Similar topics
- Methodology