If you have 3 learning strategies (Independent Variables) and some outcome measure (Dependent Variable), such as scores on a test, then ANOVA is exactly the procedure to use to determine whether there are significant differences across groups. If you obtained null results that simply means that there are no such differences.

If all you have are raw scores on measures of learning strategy, then it is pointless to compare them. A higher score would not necessarily mean that people were more likely to use that strategy; it could simply be an artifact of the questions on the scale.

Hi

Your independent variable is the teaching strategies, and it has three levels. Thus, one-way ANOVA is exactly what you need. If you are going to add another independent variable such as gender (assuming that the literature supports this inquiry), then you may conduct a two-way ANOVA.

I agree with my colleagues, ANOVA allows you to treat the results of measuring each variable.

"Is there any way to find out the significant difference among dependent variables? What test should, I use? Ima Farzana asks,

With all respect, as far as I know, there is no test to find out the significant difference among dependent variables. There a lot of tests (see previous posts) to figure out the statistically significant difference among dependent variables and independent variables.

Thank you everyone for answering. I do agree with all of you and think that there is no other test for this. But there is this past study that I have looked into who conducted paired sample t-test among the variables in looking at the differences. For eg, motivation-attitude, motivation-strategies, attitude-strategies (among science students). That is why I am quite confused.

Dear Ima,

The various forms of t-test are used when you are comparing results for two conditions. But t-tests cannot be used when you have more than two conditions. ANOVA must be used. (You can run an ANOVA with just two conditions, so it's a more general procedure.)

But it looks as though you are actually trying to compare scores on different measures, not using learning strategies as independent variables. Is that the case? Are you, for instance, trying to compare a score on a motivation scale with a score on an attitude scale? You would probably, in principle, use paired t-tests for this. BUT the results will be meaningful ONLY if the tests all yield standard scores (like T scores, for example) based on the SAME normative sample. Otherwise the results will be gibberish.

As a "for instance:" suppose I have administered an IQ test (where scores have a mean of 100 and SD of 15) and a personality test (where scores have a mean of 50 and SD of 10) to a sample of law students. I compare Verbal IQ with Extraversion score and conclude that law students are more intelligent than extraverted. That's utter nonsense! Yes, the numbers will crunch, but the results have no meaning whatsoever. The problem is more subtle if the scores are on the same scale but the norms are based on different groups, but trust me on this, it would still be nonsensical. So I'm hoping that either I'm wrong about what you're trying to do or the different scales you're using all derive from the same test.