Question on confunding variables in 2009 exam

[chickenpop]

I know that a confunding variable is when you mix up the effect of an extraneous variable for the effect of the independent variable, so conclusions can't be made. However, I'm always hopeless at these questions

I was going through the 09 exam, and the confunding variable bit got me again

Doctor Fraser is a university research psychologist. His area of expertise is the development of literacy skills
in children.

Doctor Fraser has designed a new literacy program for Grade 4 children in Victoria. It is a 30-minute television
literacy program that runs daily for four weeks.

To test this program, Doctor Fraser sent a letter to all parents/guardians of Grade 4 children in Victoria asking
for volunteers. The children of the first 100 parents/guardians who replied were accepted into the study.

Prior to the experiment, each participant sat a literacy test (Literacy Test A) administered by their Grade 4
teacher. The teachers then sent the results to Doctor Fraser.

Participants were put into 50 pairs based on gender and the similarity of their scores on the literacy test (Literacy
Test A).

A computer program was used to select, by chance, one member of each pair to undertake the literacy program.
These participants had to watch the literacy program on television for 30 minutes each day for four weeks.
The other member of the pair was allowed to watch cartoons of their choice for 30 minutes per day for four
weeks.

At the end of four weeks, the participants’ Grade 4 teachers administered a second literacy test (Literacy Test B)
and sent the results to Doctor Fraser.

Results between the two groups were then compared. A test of significance was calculated and p >0.05.
All ethical guidelines were strictly followed.



the answers doesn't say anything 'In this question students needed to name one uncontrolled variable and then explain how this variable affected results.'

How do we approach these questions?  

[Spreadbury]

I assume that because "cartoons" is undefined they could watch a variety as long as it's a cartoon. thus an extraneous variable could have been the content of the cartoons, some may have been the content of the cartoons. some may have been more educational while others were more mindless. this variable was not addressed in this question and may have been confused with the results. if one child watched a cartoon with slightly more educational content, then that person could potentially- though it is unlikely, perform better than a participant who watched a more 'mindless' or uneducational cartoon

[sillysmile]

Yes, well spreadbury has quite a good idea, also the sampling was not random, but rather derived from volunteers, and the letters that were sent would have reached and been replied to at various speeds in different regions of victoria. The placebo and experimenter effects could have possibly influenced the dependant variable. I believe there is also a suggestion that the literacy tests were of each teachers creation, which implies different tests per teacher.

[chickenpop]

ahh, i see, thanks guys