I have to admit that for awhile, I was really hesitant about the idea of passing out a survey as a possible research method; I just wasn't sure that I could truly obtain an in-depth look at student attitudes and motivations in Chemistry, and more importantly how various instructional methods influence those attitudes. However, I recently read an article that used a Likert scale to assess student self-efficacy, learning strategies, attitude towards the importance of science, etc. and this is exactly the type of information that I was interested in looking at. I think that by combining these elements with additional sections about perceived teacher attitudes and instructional methods, I can gain a lot of useful information about what students think of various elements of their Chemistry classes.
Another consideration would be who to send these surveys to, and would students even bother to respond? Would class level (AP, regular) skew any possible correlation between student attitudes and instructional methods? For example, AP Chemistry students are likely to be intrinsically motivated, regardless of the quality of instruction. And what about school demographics (ie, Reagan versus Atkins)?. I am hoping that by accounting for a variety of factors in my questionnaire (such as self-efficacy), I can eliminate some of that possible error/bias in the results.
Is this "meaty" enough of a study? Is that even a word? I really want to do something multifaceted, where I can study correlations between multiple variables. Basically I want to make things more complicated for myself... :)
2 comments:
You can control for some variables in your design study and you can conduct various statistical measures depending on your study to measure between and within-subject or even repeated variables if you give the survey multiple times to the same participants. You are wise to be concerned about possible number of responses as this will effect power.
Using a survey that has been previously established as being reliable will control for error/bias in the results.
Who to send your survey to and how to analyze your data depends on your research question. Are you interested in comparing AP and regular classes? Are you only interested in one? Are you interested in a particular school community (urban, rural, suburban? You can comment on student demographics in your participants section. If you suspect any variables that were not controlled for influenced your results, we'll mention these in the discussion section of the paper.
Casie, this is sounding better and better. If you can locate and get permission to use the prior survey, that will make your life easier. You might get a chemistry teacher at several schools to administer the survey to their classes. That would get a decent sample size.
Post a Comment