Minimising Practice Effects..

PsychChick

Hi all,

I am doing a project in which I am using questionnaires. I would like to use the questionnaire twice but giving different instructions on each. However I am worried that practice effects will affect my results..

I am just wondering is there anyway around this, I would really appreciate your advice or comments!

Thanks!:)

krankykitty

would it be possible to have a large break (days, weeks, months) between the first round of test and second?

Valmont

What sort of questionnaire is it? If it's an attitude scale of some sort then it shouldn't be a problem but if it's a test of ability then yes, you will have to try and circumvent the problem.

PsychChick

Hi,

Thanks..
Its a scale measuring perceived actual social support and perceived desired social support. I am using the MSPSS questionnaire twice with different instructions.. Im worried people with just end up answering both the same..

Valmont

PsychChick wrote: »

Hi,

Thanks..
Its a scale measuring perceived actual social support and perceived desired social support. I am using the MSPSS questionnaire twice with different instructions.. Im worried people with just end up answering both the same..

Are the instructions radically different? What you could do is give every second participant questionnaire two- first and every other participant questionnaire one first. This would counter-balance the effect that one set of instructions may have on the second. Although without more information we can't really help you.

epictetus2009

suggest that you review previous research and build on that or at least get design answers to begin with.

sounds seriously flawed from a design perspective

might i suggest that you use a different survey which looks at the same type of construct i.e. psychological well being..for part 2 of the same exp group. u may even get some build on assocation meaningful --> psycholgoical well being or vice versa..

• Psychological Well-Being Scales
• GHQ 12 scale
• satisfaction in life scale

if you need further info pm me i can supply the scales

regards

C

hotspur

What is your purpose in giving different instructions and what are your instructions which differ from the existing one?

epictetus2009

same test in short period of time is a confounding var and therefore the design is flawed but interested to get the full picture here!

New.White.Socks

It is not by any means uncommon to use the same questionnaire twice in research. Often times practice effects are not even acknowledged by researchers because it's a bit of an awkward one.
Different questionnaires is a nice idea, but it would lead to issues about what caused potential changes in the dv.
A different set of instructions is likely to help to a certain extent, changing the environmental context (instructions being a component of) means that people are less likely to just fall into the same answering style. Another thing you could do is organise a little buy-in excercise in an attempt to ensure that participants regard it as important, and genuinely think about their answers the second time as well as the first.

I don't think you will come up with a 'solution' to your problem, if you do please share!! The only thing you can do is try to minimise it and address the issue. Make it clear that you've thought about it, made an effort to minimise its effects, but still acknowledge it. It's always better to point shortcomings in your research.

Also, this is likely to result in a Type II error, which is much better. It means that if you find significant results, you've got the despite practice effects, not because of them.

JuliusCaesar

We often use the same questionnaires in before-and-after treatment; sometimes we use the same questionnaire every session.
Also, in clinical work I use a questionnaire for measuring self-esteem which uses the same questions (in different order) to measure myself-as-I-am, myself-as-others-see-me, myself-as-I'd-like-to-be (a self-discrepency measure) and as long as I explain it properly, clients don't seem to have a problem with it.