# TS-pooled or TS-unpooled?

#### moutazgendia

##### New Member
Should I used the TS-pooled or TS-unpooled to calculate the below question;

Two competing coffee shops wanted to know how satisfied their customers were with their drink choices (on a scale of 1 to 7, where 1 is very unsatisfied and 7 is very satisfied). They’re assuming that there is some difference between the two companies’ customers, but not who will be better. They each took a random sample of their customers to get a average rating.

Company A had 58 participants, who responded with an average rating of 4.52 and a standard deviation of 0.72. Company B had 62 participants, who responded with an average rating of 5.21 and a standard deviation of 1.35.

Assume that sample1 comes from Company A and sample2 comes from Company B.

Compute the 90% confidence interval. Round to the fourth decimal point. Please enter you answer in the following format: (lower_value, upper_value)

#### hlsmith

##### Less is more. Stay pure. Stay poor.
What does TS stand for? T-statistic?

#### hlsmith

##### Less is more. Stay pure. Stay poor.
The above question doesnt say anything about test, says calc CI.

#### moutazgendia

##### New Member
@hlsmith I mean should I use the formulas for the pooled or un-pooled test statistics (ex. pooled standard error or the un-pooled standard error) ?

#### staassis

##### Active Member
Run Levene's test for equality of variances. Depending on its verdict, perform either "pooled" or "non-pooled" calculation... This advice is valid if the professor has taught you Levene's test. If not, check with the professor regarding the "dummied down" approach that he/she advocates.

#### moutazgendia

##### New Member
@staassis unfortunately I don't know levene's test, I will google it and see if I can figure it out.

@GretaGarbo The standard deviations are not the same, but I can't really tell if they are similar. is there a rule for it? or agreed upon approximation?