linear regression can it be used for one group?

Can i study correlation and linear regression for only one group that answered my questionnaire? the two variables of my research are "gamification" and "motivation" but how can i study the relationship between those two without the need to give the questionnaire to two groups? (most of questions in the questionnaire concern motivation only)

• Welcome to Cross Validated! Why do you think you need two groups? // If you do not ask about gamification, unless you know some relationship between motivation and gamification, it will be tough to say much about gamification. However, that does not sound like a second group, just some additional questions.
– Dave
Mar 20 at 1:47
• Thank you Dave, because i was searching and i found that its impossible to do correlation by only checking motivation (for exemple i need to give the questionnaire to a group before gamification then again give them questionnaire again after gamification then its possible to do correlation and regression by gamification) , it is hard to find one group and give it two questionnaires before and after. i am not sure if it will be good to continue the research by only studying "motivation of students using gamification" with only one questionnaire. Mar 20 at 12:00
• Can you please clarify your research design? I first thought you want to measure "gamification" and "motivation" and then compute the correlation between these 2 measures. However, now it sound like you want to see whether the level of "motivation" changes from before to after some kind of even called "gamification". Is the latter correct? Mar 20 at 12:03
• yes exactly my research is about "the impact of gamification on students' motivation" and when i searched i found that i need to do the correlation and linear regression to study the relationship between the two, so later i found out that its impossible without having the variable of "gamification" and without giving the questionnaire before and after integrating the gamification thing. Because if i just sent the questionnaire once to a group asking about their motivatio nafter using gamification i will then need to change the title to "level of motivation..." Mar 20 at 13:39

OK based on your last comment, you have two somewhat different things going on here: (1) research design and (2) choice of analysis. In order to investigate the effect of gamification on motivation, your research design might be for instance the following:

-2 groups of individuals

-You measure their motivation (time 1)

-one group goes through gamification, one does not (the latter maybe goes through some event that is otherwise similar, but lacks the "gamification" element)

-you measure their motivation again (time 2)

Choice of analysis is a different issue. If you chose to use some version of the above suggested design, you could use multilevel regression or repeated measures ANOVA with time x group interaction predicting motivation. Correlation does not seem to be a relevant analysis with your research questions.

If you only have data on some group's motivation after they went through a gamification event, you can't say anything about the possible effects of gamification on motivation, and actually not run any analyses either (at least no correlations or regressions, if you only have one variable measured once).

• Thanks alot for your answer, my research question was "what is the impact of gamification...." but now since i will study one group to test their motivation i think the question will be "what is the level of motivation...?" , i think it is hard tho to do the study the way you said if the questionnaire is shared randomly online. Mar 21 at 10:18
• Yeah, I totally get that - we can't always run the study we'd like and need. Unfortunately, if you only measured motivation after the gamification you can't really say anything about the group's motivation level related to gamification. But, if you measured something else than motivation, you can at least compute correlations between your motivation variable and these other variables. Mar 21 at 17:35
• yes indeed! hmm maybe age and gender can be in correlation with motivation i will search for it more and thank you for your answers ^^ Mar 21 at 23:22