How to measure long-term attitude change?

[Marissa4usa]

---

Members do not see this ad.

Hi all,
I'm looking for help to come up with a construct that my prof has asked me to develop. I do not know whether a question like this is acceptable, I am not looking for the easy way out, but I am on the verge of getting really (!!) frustrated.

Okay, so...the problem is that my prof has asked me to come up with a construct to assess change in attitude/outlook on a certain aspect. I don't feel completely comfortable sharing it publicly because it is my prof's research but if someone would like more info, feel free to send me a PM.

Anyways, he wants me to come up with a way to measure how the attitude of a certain population changes over time or rather, what the factor/variable is that contributes to the change. Problem is that can't really be a within subject design because this change in attitude occurs occurs over a time-span of several years and our subjects are college students (which are actually the target population) and we don't have the means to pay people to commit to a larger study anyways. So, research suggest that this change in attitude occurs sometime when people are in college (or around that age in general) but it isn't necessarily the result of persuasion or something along those lines. It's something that "just" occurs and we're not sure what it exactly the variable is that this attitude change can be attributed to. We have some guessed based on the existing literature but pretty much this will be an exploratory study because not much research has been done in this area.

I pretty much feel like I have to come up with the impossible. It's not so much the potential factors that we think that might be responsible for the attitude change but how to use data from one time-point to make the inference that certain variables are responsible for the change.

Maybe I am missing something that is really obvious but ANY input would be appreciated
ME --> 😕 😕 😕 😕

 

[KillerDiller]

My first thought is that you may need to construct some sort of interview protocol to get at your research question. If the change in attitude "just happens" then it may take a rather in depth conversation in order to get at the individuals' experience of that change.

Have you tried looking at adapting something like Marcia's Identity Status Interview? It captures change retrospectively by examining whether the individual has dealt with internal conflict.

 

[Marissa4usa]

My first thought is that you may need to construct some sort of interview protocol to get at your research question. If the change in attitude "just happens" then it may take a rather in depth conversation in order to get at the individuals' experience of that change.

Personally, I would LOVE to collect qualitative data on that but my profs wants quantitative data. At the very most he'd be willing to do some content analyses of some sort of short paragraph they write.

Have you tried looking at adapting something like Marcia's Identity Status Interview? It captures change retrospectively by examining whether the individual has dealt with internal conflict.

This looks interesting but I don't know whether it would work for my prof.

 

[justme08]

Doubt, I'm much help, but it sounds like one of the things your prof is trying to pin down is an avg. age or time of this change. So, if the prof is sure it happens in college couldn't you do something where you collect data about their belief/attitude towards X, then two months later, then four months later... or, perhaps could you use PDAs and quiz them periodically about X to see if they've changed?

 

[Marissa4usa]

Doubt, I'm much help, but it sounds like one of the things your prof is trying to pin down is an avg. age or time of this change. So, if the prof is sure it happens in college couldn't you do something where you collect data about their belief/attitude towards X, then two months later, then four months later... or, perhaps could you use PDAs and quiz them periodically about X to see if they've changed?

It's not really about age or time (like I said, I know what the study is on, I just don't feel comfortable sharing his research ideas on the net)

It's difficult to explain, it's not something occurs within a few months but it's a very gradual process from adolescence into adulthood that these attitude and hence one's behavior changes. There is plenty of research that has looked at people's attitude at all ages, but it's not really clear what leads to the change, (i.e. what the variables are. At this point it could be anything from simply maturing to *whoknows*. It's a rather exploratory project, so we don't have any too specific hypotheses that could lead us to a certain construct.

And again, my program is located in a pretty small school, so we don't really have the funds to run longitudinal studies :-(

 

[justme08]

Maybe try it with a cross sectional study. Then you could see perhaps specifically what age has what you are looking for then you could perhaps target your study there on the people that already have said change?

 

[KillerDiller]

Do you have a way to measure current attitudes toward the issue in question? If so, you could have participants fill it out twice--once by responding with how they currently feel, and once with the instructions to complete it with how they remember feeling two years ago (or however long).

 

[Marissa4usa]

Do you have a way to measure current attitudes toward the issue in question? If so, you could have participants fill it out twice--once by responding with how they currently feel, and once with the instructions to complete it with how they remember feeling two years ago (or however long).

That would probably a very easy thing to do. However, it is not really assessing what my prof is trying to find. It's not that we don't know that there is a change -like I said the attitude on this issue has been researched with all kinds of populations and the evidence clearly suggest that people's attitude change on this issue - it is more that we want to find out (1.) what brings the change about and (2.) how we can measure it using only data from one time point.

My brain is really fried 😱

 

[Marissa4usa]

Maybe try it with a cross sectional study. Then you could see perhaps specifically what age has what you are looking for then you could perhaps target your study there on the people that already have said change?

Unfortunately, we are limited to college students and in this case there won't be too much of a difference between, say, freshmen and seniors. This kind of change is not really clear cut. It's more or less something along the lines of outlook-on-life - kinda-thing (i.e. how you deal with certain things).



I know not being able to specifically say what this study on certainly doesn't help and my prof isn't really clear on what he wants either, but I truly appreciate anybodies input.

 

[docma]

Could you find a way to sample now and then match with some element of your variable that could be identified in admissions essays--thereby getting a time 1 from the application year, time 2 from current measure and then do interviews to see what emerges from the participant re their construction of the "change"? Sociologists have a much wider catalog of interesting ways to gather data so I'd look to some of their study designs as well

 

[Marissa4usa]

Could you find a way to sample now and then match with some element of your variable that could be identified in admissions essays--thereby getting a time 1 from the application year, time 2 from current measure and then do interviews to see what emerges from the participant re their construction of the "change"? Sociologists have a much wider catalog of interesting ways to gather data so I'd look to some of their study designs as well

Unfortunately, what we are looking at, does not come up even remotely in admission essay and if so, it would be very weird 🙄.

 

[docma]

You'd be surprised what comes up.....however, I can see how many topics would be a stretch. The point is to consider ways in which "time one" data may be archival somewhere