HW 8, Problem 1

[chelsi]

I am so lost as to start. On question two in problem 1:

"Choose an observation in the “group” sample, count the number of observations in the
“singly” sample that are smaller in value; count 1/2 for each tied observation. Repeat this
for all five observations in the “group” sample and add the numbers up to obtain K1 and
report the value of K1. (Note that a large K1 is the evidence that the distribution of the
“group” is greater than the distribution of the “singly”.)"

What do you mean count half for each tied observation??? This is what I have:
group sample observation: 3.2, 3.4, 4.0, 4.1, 4.8
# of obersv. in singly smaller in value: 3, 4, 5, 5, 5

but what do you mean by half for each tied observation? Instead of 5 in my number of observations smaller than would I have 5.5, 5.5, 5.5????

[Taeho Kim]

chelsi wrote

What do you mean count half for each tied observation??? This is what I have:
group sample observation: 3.2, 3.4, 4.0, 4.1, 4.8
# of obersv. in singly smaller in value: 3, 4, 5, 5, 5

This looks just good for the sub-question #2.

The "tie" means the case when we have the same values in the first and second sample.
In this example, we don't have any "tie's", so you could ignore that part.

The example in the page 15 of Unit 14 slide should be an example of the case with "tie's".

[chelsi]

Ok great. That's what I thought, but I wasn't 100% sure. Thank you!

[Taeho Kim]

Ok great. That's what I thought, but I wasn't 100% sure. Thank you!

chelsi wrote

Ok great. That's what I thought, but I wasn't 100% sure. Thank you!

Sounds good, chelsi!