I am doing a study on the presence and prevalence of symptoms of Autism in children with ADHD. I am not sure what statistical tests to use to analyse it and would really like some help! I think my interest areas are....
- the presence of symptoms in the sample
- the prevalence of symptoms in the sample
- the effect of ADHD severity on symptoms (correlation?)
- the effect of quality of life on the presence of the symptoms
- the effects of the variables of age, age at diagnosis, gender, medication use, siblings, and siblings having a learning disability on the presence of symptoms (multiple regression?)
I would really appreciate if you were able to help me with this!! Very confused and getting very stressed about it!
I am not quite sure I understand what stage you are at. Have you obtained your data yet? Which statistical tests are appropriate are a function of the type of data which you have.
I really recommend working through a very simple text book such as Learning to Use Statistical Tests in Psychology (I can email it to you if you wish and pm me your email address). Mind you, there are now innumerable sites which offer very good guidance.
To my mind you simply have to understand the basics of statistical tests if you doing research, otherwise you will continually feel like you currently feel even if you get advice about what tests you should be doing. Bottom up is the way to go with stats and research in my opinion.
Looking at comparable research is then usually a good indication of the appropriateness of which tests to use if you are unsure.
Andy Field's Discovering Statistics series (for SPSS, SAS, and an 'R' edition coming out this month) is well regarded - and one I've found particularly useful. Well written with plenty of psychology-oriented examples. As above, invest some time now familiarising yourself with the principles of inference - it is much easier to approach research design and statistics with intuitive understanding. Field is particularly good at explaining multiple regression.
Definitely recommend Andy Field, he's probably the best person at explaining statistics.
Is it just one group? Or are you comparing two groups- one ADHD, one non ADHD?
Are the autism symptoms your outcome variable? A multiple regression will show you how much of the symptoms are accounted for by each of the variables you have mentioned.
In other words, how much the autism symptoms are attributable to age, how much are attributable to medication use etc.
I think a little more detail is needed.Would i be right in saying that you would start with your hypotheses, eg. learning disablity such as Dyslexia/Dysgraphia has a significant affect on the maintenance of symptoms of adhd, this would however be quite hard to carry out, itd be long term as the prevalence of adhd would have to be measured at different intervals and even then you would you have to take various measure to suggest that its the learning disorder having an affect on adhd.
youd have to start by quantifying each of these concepts eg. ones quality of life, age at diagnosis, ADHD severity, symptoms...... using verfiable or standardised tests/surveys/questionaires and then placing them against eachother in correlations/ regressions etc...to essentially suggest theres a relationship between them.
if you are not intending on quantifying any of these though its qualitative research you're lookin at, and in that case spss is no use.