Psychology AQA A PSYA4 18th June 2012!

like how to remember which stats test to use when? and the justification for using them?

xx

For people who are struggling.......

We need to know 4 tests (obviously lol)
Mann-Whitney
Chi-Squared
Spearmans rho
Wilcoxon

You use them when.......

Mann Whitney: A test of difference, independent groups, ordinal data
Spearmans rho: Test of correlation, ordinal data
Chi-sqaured: test of association, independent groups, nominal data
Wilcoxon: tets of difference, related groups, interval data

In the exam they will give you an observed value...in order to determine whether this observed value is significant it is compared to critical values and these are given in a table for that statistical test
Before using the table you need to know whether you had a directional hypothesis (one tailed test) or a non directional hypothesis (two tailed test) and you need to know the number of participants used (N).

For the Wilcoxon test and the Mann Whitney test the OBSERVED VALUE must be EQUAL to or LESS THAN the CRITICAL VALUE for SIGNIFICANCE to be shown

For the Chi-squared test and Spearman's Rho the OBSERVED VALUE must be EQUAL to or GREATER THAN the critical value for SIGNIFICANCE to be shown.


I hope that doesn't sound confusing... its really simple to understand (:

Thanks.Whats independant groups, mathced pairsand...forogt the other one

Independent: Participants are randomly allocated to either one or other of the conditions

Repeated: The same participants are used in both conditions.

Matched : participants are matched as closely as possible with another participant and then the pairs are randomly allocated to either one or the other conditions

(cheers, just need someone to revise with you see cos i'm dumb.., just add stuff in,and tell me if i'm right or not..)

Independent: <---disadvantage - individiual differences, e.g. one group may consist of people with higher IQ in totoal, compared to the second group.

Repeated: <---disadvantage- may have order affects, they may perform better the second time round.

Matched : participants are matched as closely as possible with another participant and then the pairs are randomly allocated to either one or the other conditions <----(isnt this rong )

i thought matched pairs was when there are like
4 groups:: group A, group B, group C, group D.
and 2 conditions.
wait...i'm not sure my self what i was going to say
explain more please, and what the disadvantages are

yeah like how do you figure out the one you're supoosed to use, with the datra given?

GOOD NEWS!!

COMPARE & CONTRAST QUESTIONS WILL NOT BE USED:

I asked Cara Flanagan on her blog (the writer of complete companion books) and she responded yesterday with this:

"Hello

Compare and contrast questions won’t be used – they were used before the specifications were revised 3 years ago but have not appeared since. They require you to consider the similarities and differences between two things, and for AO2 to evaluate these e.g. look at evidence, implications etc.

Cara"

Link:
http://www.oxfordschoolblogs.co.uk/psychcompanion/blog/community/topic/psya4-help/page/2/

no your first explanation was right - matched pairs is when you basically "match pairs of people" - you have about two conditions to test from which is usually the experimental condition group and a control group to compare from.

Conditions A, B, C and D etc would still be independent group measures if each condition is different but you have not specifically tried to match people on similar characteristics I believe; hence they are "independent" of one another.

Repeated is simply when you use one group of people and have them try all the conditions one at a time - hence "repeated".

hope this helps? its how I remember.

GOOD NEWS!!

COMPARE & CONTRAST QUESTIONS WILL NOT BE USED:

I asked Cara Flanagan on her blog (the writer of complete companion books) and she responded yesterday with this:

"Hello

Compare and contrast questions won’t be used – they were used before the specifications were revised 3 years ago but have not appeared since. They require you to consider the similarities and differences between two things, and for AO2 to evaluate these e.g. look at evidence, implications etc.

Cara"

Link:
http://www.oxfordschoolblogs.co.uk/psychcompanion/blog/community/topic/psya4-help/page/2/

I've got a weird way of remembering which test to use when.
4 sentences -

Sophie Can't Read Indexes
(Spearmans Rho) (Correlational) (Repeated measures) (Interval data)

Men Differ In Intelligence
(Mann-Whitney U) (Difference) (Independent groups) (Interval data)

Walter dives regularly In Ocean
(Wilcoxan T) (Difference) (Repeated measures) (Interval/Ordinal data)

Chris DAnces Independently Normally
(Chi squared) (Difference/Association) (Independent groups) (Nominal data)

Obviously, chop and change the words in the sentence as much as you want, but I find the more obscure the sentence the better it stays in your head. So now whenever I get a question asking me to name which test to use I just write out these four sentences and work it out from there
Hope it makes things easier to remember

Does anyone have any predictions for media?

Ehh, I still don;t get it
Isn't it like this.
group A and B > Group A goes to condition 1,then condition 2, group b go to condition 2, then condition 1.
group C and D> group C goes to condition 1, then condition 2, group D go to condition 2 , then condition 1..

but how do you say that in laymens terms
all what are the disadvantages, I know the advantage is that, order affects are less likely in this desgin.

This is called counter-balancing. It's the way to get around order-effects in the repeated measures design. Matched Pairs is how a couple of people have explained it above: the researcher matches similar people and then sends them to the different conditions, hence seeing how much of an effect the IV has on the DV.

Sure, I meant to say how do you sample if one of the variables in the correlation was to do with the participant themselves, e.g. correlating participant age with scores on an IQ test or something similar, because obviously you need a range of different ages.

Lol,you've actually explained it well. Thankyou for clearing that!
My teachers always told me this is what matched pairs is

Can you please go through what Wilcoxton T test basicallly means in laymens terms as well

The Wilcoxon test is used for a test of:
- Difference
- Repeated Measures
- Ordinal Data (and Interval, I think).

That's all you need to know, really!

But in ma book there's a whole 2 pages on this!
And please explain to me what 'difference, and associated' is. thankyou