Need statistics help... or is it sadistics?

[Sounder]

I know, this is totally random... but please do me a favor and not hijack the crap out of this thread until I get a good answer.

I need a little help on correlation versus causation. I get the concept - easy breezy. My prof wants us to construct some stupid triangle and when I try to ask him questions, he doesn't really answer them... and there is a serious language barrier (he can't hardly speak English, and I don't speak Korean).

So either I need someone who can speak Korean to talk to my prof, or perhaps someone here can help me understand which two variables are dependent and which third variable is independent.

Here's my example:

Statistics show that fat people buy more diet food than skinny people. Therefore, diet food must cause people to get fat.

I think the two dependent variables are fat people and diet food, and the independent variable is that diet food makes them fat.

OR

Are the two dependent variables "fat people are fat" and "fat people buy diet food" and the independent variable is the "diet food?"

I'm just very confused by my non-English speaking 30+ tenured professor's lack of explanation of a concept that will be on the final exam this Thursday. :shootself: He's never received a stellar student review yet it's mandatory that everyone take his class. I think he's still wearing the same pants he wore when he started teaching - they're WAY late 60's early 70's polyester suit pants with no belt loops, no buttons, 1 pocket, and no suspenders.

Anyone help me out here?? The triangle looks has the IV at the top, and the DV1 & DV2 at the bottom (not that it really matters, but maybe it's a popular concept that out-of-touch sadistics professors who don't speak English teach).

Sigh.

[dsteding]

You are doing it wrong.

[Sounder]

You are doing it wrong.

I hate you.

[cardiver]

You are doing it wrong.

I hate you.

I thought that dsteding was refering to your comment, "20 minutes my ass"........

[Sounder]

You are doing it wrong.

I hate you.

I thought that dsteding was refering to your comment, "20 minutes my ass"........

I hate you too.

[boydski]

Statistics show that fat people buy more diet food than skinny people. Therefore, diet food must cause people to get fat.

Hi Doug,

I can't really help you with your Statistics, but from a Logic point of view, the above statement fails the rules of Logic by being "Post propter hoc, ergo propter hoc". Which is Latin that roughly tranlates as "after, therefore because of".

Don't you love it when we muddy the waters....

[dsteding]

Statistics show that fat people buy more diet food than skinny people. Therefore, diet food must cause people to get fat.

Hi Doug,

I can't really help you with your Statistics, but from a Logic point of view, the above statement fails the rules of Logic by being "Post propter hoc, ergo propter hoc". Which is Latin that roughly tranlates as "after, therefore because of".

Don't you love it when we muddy the waters....

Hmmm, one of Sounder's statements failing the rules of logic? Shocker.

[edit] maybe if you spent more time doing homework and less time on the boards, you'd understand your teacher. As a former TA, I'm having too much fun with this.

[Sounder]

Yeah, perhaps if I spent less time on NWDC I could learn Korean. Thanks for the tip!

Yes, the logic does fail (that's the point of the exercise) - we have to make a statement which shows an example of where correlation does NOT mean causation... and we get extra points for humor. This was the best I could do... but if any of you have a funnier idea, I'd love to hear it!!

Oh yeah, and if anyone actually knows the IV, DV1, and DV2, that'd be awesome too.

[LCF]

I was never good at statistics, but I majored in math.

In general, independent variables are the things which define the problem; dependent variables are the things that result from applying the function to the independent variables. In this case, I would say a population and its diet food consumption are the independent variables, and the prevalence of obesity is the dependent variable, once you apply the statistical function that combines a and b and produces c.

[Pez7378]

When is it okay for me to Hijack this thread? This crap gives me a headache. ](*,)

[Kalatin]

Doug,
As a caveat, it has been about 7 years since my last statistics work. The problem is that the problem presented does not really fit into IV/DV distinction and doesn't make any rational sense to me. There are many who think that IV/DV should only be applied to experimental work when it is clear what you are manipulating (the independent variable) and what you are studying (the dependent variable). That being said, I would guess that the IV is the amount of diet food purchased. The two DVs would be the resulting category of subject (fat vs. skinny).
In an experimental sense, you would control the amount of diet food a subject purchased (the IV) and then determine what category they fell into at the conclusion of the experiment (fat or skinny). Stupid experiment, but it seems to fit the problem given. It also illustrates the general problem of correlation vs. causation.
Make sense?

[Sounder]

When is it okay for me to Hijack this thread? This crap gives me a headache. ](*,)

I'll let you know... but not yet. I'm studying my Korean... and Lynne and Kate are helping in English.

[Kalatin]

I would listen to Lynne. She does math. I did statistics as a part of experimental psychology - different beasty.

[Sounder]

Ok, I can see what Lynne's saying but I actually think Kate is closer to the Korean translation of what he wants.

DV1 = Fat
DV2 = Skinny
IV = Volume of diet food purchased

This fits the dude's triangle better than anything I've come up with thus far.

In other news, I emailed him - I'm REALLY hoping he types in Korean and then allows a translation program to put it into English. I'm hoping that I will get a more clear answer in-writing than verbally.

The rest of the final is tough, but I've got all the complex and difficult concepts nailed down - his damn pet-problem triangle monkey-humping-a-football thing (it's almost yours Pez, just give me a bit longer to ensure I've got it) is what's absolutely killing me.

[Huskychemist]

Not sure about the triangle thing...

From a science perspective, I would argue that the amount of diet food consumed is the IV, and some measure of obesity (e.g. BMI) would be the DV.

And remember, correlation does not prove causation.

[Paulicarp]

since I can't sleep and I can't dive, I might as well take a stab at this one.

Independent variables answer the question "What do I change?"
Dependent variables answer the question "What do I observe?"
(according to the infalible wisdom of wikipedia)

sooo...I would think that means that consumption of diet food is the IV because it's the thing that you choose to change between the two observed groups and the rate of obesity is the DV because that's the result you are looking to observe in each group. Obesity in diet food non-eaters is DV1, Obesity in diet food eaters is DV2, and eating diet food is the IV. In this case we observe a strong corelation between the IV and DV2.

I drive around looking for severely narced people and I ask them what they do for fun, and invariably they say "diving". All the non-narced people say they only dream about diving when they get certified. Therefore severe narcosis is corellated with diving. Narcosis is the IV because it's the thing I changed between the two groups of people. You observe the results between the two different groups and see that diving is the DV1 and dreaming about diving when certified would be DV2.

...now where was I...oh yeah...p1+p2+p3+p(n)=P

[Grateful Diver]

I believe that logic triangle looks something like this, Doug ...



... Bob (Grateful Diver)

[dsteding]

That is funny. Will someone please hijack now?

You know you've gone too far on scuba diving boards when you seek advice on your homework, I'm just sayin'.

I believe that logic triangle looks something like this, Doug ...



... Bob (Grateful Diver)

[Sounder]

I'll admit, it is pretty pathetic.

I am however always impressed with the diversity of knowledge here though, so I figured it'd be worth a try.

Unfortunately, his "triangle" nonsense is not in the text at all - all verbal lecture and quite unclear despite everyone's questions which only seemed to confuse him.

Oh yeah, and I paid something like $2500 for the "opportunity" to take this required course... ugh.

(crossing fingers my prof can write English better than he can speak it)

Thank you all for your help... even you, Steding.

John - thanks for the PM... I've got some notes and I think that may be all the Korean I need. :angry:

Chris, you've got the green light. Make it a good one.

[vbcoachchris]

Simple answer

Dependent Variable is what is measured
Independent Variable is what is change in the experment (or situation)

Scott

[vbcoachchris]

Simple answer

Dependent Variable is "what is being measured?"
Independent Variable is "what is changed in the experment (or situation)?"

Scott

[dsteding]

Thank you all for your help... even you, Steding.

My pleasure, glad to be of help.

[Pez7378]

Statistics show that more fat people buy diet food than skinny people, therefore diet food makes people fat. Skinny people exercise more than fat people, therefore exercise makes people skinny. Furthermore, diet food is expensive and since more fat people buy diet food than skinny people, fat people are statistically richer than skinny people. Since skinny people are not as rich as fat people, and cannot afford to buy diet food, skinny people don't get fat. Since diet food tastes like crap, the rich fat people that buy it, don't actually eat it and buy a Supersized Burger Meal instead.

Sadistically speaking, picking on fat kids in the school yard, does not make them skinny, but it does make them angry. Sadistics show that More fat kids than skinny kids beat up school yard bullies. This indicates that bullies are weak, skinny kids are weaker and you shouldn't pick on fat kids.

I hope this helps.

[Sounder]

Statistics show that more fat people buy diet food than skinny people, therefore diet food makes people fat. Skinny people exercise more than fat people, therefore exercise makes people skinny. Furthermore, diet food is expensive and since more fat people buy diet food than skinny people, fat people are statistically richer than skinny people. Since skinny people are not as rich as fat people, and cannot afford to buy diet food, skinny people don't get fat. Since diet food tastes like crap, the rich fat people that buy it, don't actually eat it and buy a Supersized Burger Meal instead.

Sadistically speaking, picking on fat kids in the school yard, does not make them skinny, but it does make them angry. Sadistics show that More fat kids than skinny kids beat up school yard bullies. This indicates that bullies are weak, skinny kids are weaker and you shouldn't pick on fat kids.

I hope this helps.

Chris, you're actually closer to getting it than you think!