4) Assume that Var. 1 is a person’s test score on a test that you
developed to predict the clinical presence of obsessive-compulsive
disorder. Var. 2 is the person’s score on the Beck Depression
Inventory. Var. 3 is the person’s score on the Gold Standard test that
has been used in clinical settings for the last 30 years to diagnosis
obsessive-compulsive disorder. Listed below are the properties of Var
1.

Mean 7, SD 1.5 r = .85 

a. As the developer of Var. 1 are you happy with your results. Explain
why and be sure to discuss your answer in terms evidence of validity.

b. On your test the clinical cut off for being diagnosed with
Obsessive-compulsive disorder is a score greater then one standard
deviation above the mean. An individual obtains a score of on your
test. Does that person have obsessive-compulsive disorder? Why?


Var1: 5 6 9 4 8 

Var 2: 23 25 23 27 26 

Var 3: 14 17 21 13 19 

=================There is also some tips for #4 :

) For the last question use the mean, SD, and r provided.  Do not 
calculate it.  However, you will have to use the data provided to 
compute correlation's to answer the remaining part of the question.

2)  The obtained score for the last part of the question should be 8.

3)  Evidence of validity refers to validity, not reliability.  These
are
two different concepts. 
===================

I tried this, but I am getting nowhere with this! 

******************************
It would be great to have the ANSWER BACK BY 3pm SATURDAY PST. After
that, the answer is useless to me.
*****************************

thanks,
hm

1. 

a. The developer of Var. 1 should be happy with his or her results.
The mean of the distribution of scores is 7, while the standard
deviation is much smaller in comparison, meaning that most scores lie
close to the mean. There is not very much variation in terms of
average score on the test. Also, the correlation (or validity
coefficient) between the test score and presence of obsessive
complusive disorder is .85. A perfect linear fit is 1.0 and two
unrelated variables have a corrleation cooeficient with an absolute
value of zero. A correlation of .85 is quite strong and implies that
the test is a good predictor of the disorder. R2, the correlation of
determination, is 72% which also implies a large majority of variation
in the test scores is explained by the line that relates the variables
of test scores and disease.

b. If a test taker achieves a score of one (1) on the test, then they
are (7/1.5) or 4.7 standard deviations below the mean. A person with a
disorder is said to be 1 standard deviation above the mean, so this
test taker definitely does not meet the cutoff to be considered
affected by the disorder.

---

Request for Answer Clarification by hotmark-ga on 21 Jun 2003 21:24 PDT

I need all math calculations for these problems please, not just the explanation.

---

Request for Answer Clarification by hotmark-ga on 21 Jun 2003 21:55 PDT

You may want to keep this in mind, while you think about my clarification request:

1) For the question use the mean, SD, and r provided.  

Do not calculate it.  However, you will have to use the data provided to 
compute correlation's to answer the remaining part of the question.

2)  The obtained score for the last part of the question should be 8.

3)  Evidence of validity refers to validity, not reliability.  These are 
two different concepts.

---

Clarification of Answer by chis-ga on 23 Jun 2003 09:24 PDT

The answer I made earlier takes into account the hints that you
provided (except for the reference to the number 8). In part A, the
question is why should the provider be happy with the results. The
first reason is that the standard deviation is a small compared to the
mean. To express this mathematically, (7/1.5) indicates that the mean
is over 4 times larger than the standard deviation. Also, r, the
number that determines the strength of the relationship, is .85, which
indicates a strong relationship. R2 or (r^2), is the coefficient of
reliability, and is (.85)^2 or 72%, which is a large percentage. This
indicates that most of the variation in presence of the disorder is
explained by the test score achieved. There is no further math to do,
keep in mind that statistics foucuses on analysis and the meaning
behind the numbers rather than a large amount of computation.

In part b, you said that "An individual obtains a score of on your
test. Does that person have obsessive-compulsive disorder". I assumed
that you said he scored a 1, which is above the mean. The question
states that the score most be one standard deviation below the mean,
meaning that this score clearly isn't enough to indicate a disorder.

I fully understand the question as it is stated and can further help
you. Please request further clarification if you have any questions or
wish to clarify the question. I assumed that in part b, the score
asked for was 1, but it seems to be written "on", so I'm not sure if
you missed a number or meant 1. which might explain the reference to
the number 8.

Responded well, but not on time to be useful