Psychology-related Question

[Whisper]

Ok all you anandtech psych majors out there, gotta question for ya that I'm confusing myself on:

My task is to come up with an experimental design and write it up as though I'd conducted the experiment. My basic paradigm is that I want to test the effects of bilingualism on verbal short-term memory. I came up with three different levels of bilingualism and labeled it as my IV (with verbal short-term memory as my DV). No big problems there.

However, here's where my confusion arises. I "tested" short-term memory by means of the same test, but administered verbally in one case and in a reading form in another. The main reason for this was because our TA wanted us to run ANOVA's as opposed to t-tests on our results. So, what I'm wondering is whether the type of memory test I gave is an experimental condition,or if it's actually a second independent variable. I was leaning towards the former and first, but now I'm leaning towards the latter.

Also, if it's simply an experimental condition, then can I eliminate one of the two test types from my experiment and still run an ANOVA on just one IV with three levels? I've forgotten most of what I learned in psych. statistics (heh) and don't have my notes or book down here at home with me, so I can't look up the answer...but I thought I remembered ANOVAs requiring at least two IVs, as anything with only one IV and a DV would be a t-test. But then again, it might be that t-tests are for one IV with one ortwo levels, and an ANOVA is possible when the IV has three or more levels.

So yeah, if anyone knows the answer then please get back to me. Thank yas.

 

[psydancerqt]

it sounds to me that your experiment is the differences in learning style. learn better verbally or reading? sorry, but i dont know what ANOVA is.. what's that? do i have to get my stats book out....?

 

[Looney]

Originally posted by: psydancerqt
it sounds to me that your experiment is the differences in learning style. learn better verbally or reading? sorry, but i dont know what ANOVA is.. what's that? do i have to get my stats book out....?

ANOVA is a statistical method= Analysis of Variance between groups.

So, what I'm wondering is whether the type of memory test I gave is an experimental condition,or if it's actually a second independent variable.

I would say it's an experimental condition.

 

[Spoooon]

A t-test is limited to situations where thare are only two treatments to compare. An ANOVA is for two or more treatments.

Your IV is bilingualism level? If you have three levels, you could consider that to be three total treatments. The two types of tests given (verbal and written), I would treat as different experimental conditions. In short, you would be running two experiments. I think that would simplify matters.

How are you assigning bilingualism levels?

 

[Whisper]

ANOVA stands for analysis of variance...it basically measures the effects that multiple variables have on a dependent variable. It's similar to a t-test, but on a more complicated level. At least as far as I remember...don't have anything here with me to check.

And the IV that I chose (level of bilingualism) is actually a pseudo-IV...mainly because it's not something I can actually manipulate personally or experimentally, I can only manipulate it by assigning different participants to different groups (sort of like choosing sex or hair color as an IV...they're predetermined).

 

[Whisper]

Your IV is bilingualism level? If you have three levels, you could consider that to be three total treatments. The two types of tests given (verbal and written), I would treat as different experimental conditions. In short, you would be running two experiments. I think that would simplify matters.

Yeah, I was thinking of the two experiment idea myself as well...I'm trying to remember what it was that made me think of test type as an IV rather than a condition.

In my write-up thus far, I stated that test type was my second IV, and that I would run a three way ANOVA to obtain results. So in essence, I could then just eliminate one of the test types from my experiment (verbal or written), go down to one IV with only one experimental condition, and just use a two-way ANOVA on my three levels of bilingualism? I'm mainly confusing myself on the type of ANOVA to use, heh.

As for howI assigned bilingualism levels: one level was English monolingual as a basis of comparison, one was early English-Spanish bilingualism (began learning English and Spanish simultaneously before age 5 years), and the last was late English-Spanish bilingualism (native English speaker having learned Spanish anytime after age 5 years).

 

[Spoooon]

And the IV that I chose (level of bilingualism) is actually a pseudo-IV...mainly because it's not something I can actually manipulate personally or experimentally, I can only manipulate it by assigning different participants to different groups (sort of like choosing sex or hair color as an IV...they're predetermined).

What I mean is how are you assigning the people to each group? On what basis? Also, there is the fact that familiarity will enhance short term memory. I'm guessing that people whose native language is not english and is their primary mode of communication will perform more poorly than for those who uses english much more often.

edit: got distracted by the dogs before I could hit enter.

 

[Whisper]

yep, took familiarity into account, which is why I wanted native english speakers only...or those that learned english and spanish simultaneously. Of course I can't really control for their amount of exposure to both languages completely,but that's part of what I'm trying to test anyways, so s'all good

I'm mainly just confused about the test type deal. If I remove it completely and only have one condition, then I'm wondering if I can still even run an ANOVA. I do have three levels in my one IV though, so I'm assuming I could...just don't know if it'd be two-wayor three-way.