Help! {Exam Questions)

[Glockmeister]

Thanks Glockmeister...hopefully ill get better as i do these questions..

ok heres is a vcaa question .. VCE 2002 Q16. What would Dr. Brodsky's conclusion based on the results of his experiment???(his experiment p-value was lower than 0.01)

i wrote none as they are descriptive stats but the actual answers are : "The faster the tempo of the music listened to by the participants the greater the speed limit was exceeded OR The more beats per minute in the music the greater the kms above the speed limit the person will go."

so what???can you draw conclusions from descriptive stats provided that your results are statistically significant??

The p-value is as much  part of the result of the experiment as everything else you know. In a journal article, the p-value appears in the results section.

[jinny1]

The answer given by the vcaa does not speak anything about the p-value in terms of the conclusion rather talks about the correlation between the results indicated by a descriptive statistic...
why is it doing that?

[laijiawen]

The answer given by the vcaa does not speak anything about the p-value in terms of the conclusion rather talks about the correlation between the results indicated by a descriptive statistic...
why is it doing that?

Which VCAA exam paper was that?

[jinny1]

2002 Q16

[minilunchbox]

2002 Q16

Different study design, different course.

[Glockmeister]

The answer given by the vcaa does not speak anything about the p-value in terms of the conclusion rather talks about the correlation between the results indicated by a descriptive statistic...
why is it doing that?

Oh.

That's correlation which is a completely different concept from the experimental methods. Correlation was removed at the 2005 revision to the study design (IRRC, it is coming back).

If you want to know about the different experimental designs beyond the one yuo've been taught in class, have a look at my uncompleted (sorry guys, too busy) Research Methods Guide

http://vcenotes.com/forum/index.php/topic,26827.msg286259.html#msg286259

[jinny1]

if its correlation then its great relief!

but the actual question asks what conclusion can be drawn.

maybe i just shouldnt do exams that have older designs

[Glockmeister]

I actually had a read of the question: this is it right?

Questions 15 and 16 refer to the following information
Dr Brodsky was interested in whether the speed at which people drive is related to the tempo of the music they
listen to while driving. The tempo of music is its quickness. This is measured by the number of beats per minute
(bpm). He conducted an experiment with 28 students who drove in a driving simulator under four conditions:
• condition one involved students driving without listening to music
• condition two involved students driving to music at 40–70 beats per minute (bpm)
• condition three involved students driving to music at 85–110 bpm
• condition four involved students driving to music at 120–140 bpm
The results of Dr Brodsky’s experiment are illustrated in Figure 5. When he tested his hypothesis, Dr Brodsky
found that his results were statistically significant (p < .01).

That is an experimental design (says so too). Probably should've read the question properly.

I think the reasoning that the VCAA has is that because it is significant, you can now use the descriptive statistic to help you determine the direction of the conclusion, in other words now that you've determine that there is an effect, you can safely use the mean differences between each of the categories to determine which of the categories are higher (in this case, sped faster) to make the direction of your conclusion. On that basis you could say that because each of the categories (loudness) were increasing with average kms over speed limit, and that these increases were significantly different, you could safely conclude that increased loudness is linked with increases speeding.



(Take everything I've typed above as the way you're are to do it with VCE Psychology, 'cause strictly speaking everything I have written above is completely and utterly wrong and that 71% of people who got that right have actually blown it and gotten it wrong. The answer is that you can't actually make a conclusion on the basis of the information given along. Again, the reason for this is actually beyond VCE again [yep, that thing again], but it relates to the type of statistical test one would use, and that you can only determine that some of the categories might be significant with each other and that further testing [of the statistical kind] needs to be done)

[masonnnn]

i'd/ i have avoided the old study design exams. some of the stuff has been disproven now so is actually incorrect and unless you can get a list of questions which still remain relevant a lot of it is...irrelevant.