determining if groups are similar to what's expected

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

determining if groups are similar to what's expected
Message posted by sam on November 18, 1999 at 12:00 AM (ET)

Market analysts believe that nationwide, approx. 50% of the population purchase a movie video at a video store on the spur of the moment. These analysts also believe that approx. 40% who purchase a movie video do so after thinking about the movie earlier in the day. Approx. 5% of the pop. think of the movie they purchased 1 or 2 days ahead and 5% think of the movie more than 2 days ahead of the purchase. Suppose that this survey was repeated in St.Louis with the 200 video buyers. Determine whether the responses for this group are similar to what would be expected in the nationwide survey. Use a 5% significance level.

Spur of the Moment: 91
Earlier the same day: 75
1-2 Days ahead: 18
More than 2 days ahead: 16

How can i solve this problem??? Please help a.s.a.p.!!! Thanks.

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: determining if groups are similar to what's expected
Message posted by Bill on November 18, 1999 at 12:00 AM (ET)

Use a one-way chi-square. For Spur of the Moment the expected frequency is 50% of 200, Earlier the same is 50% of 200, 1-2 days ahead is 5% of 200, and >2 days is 5% 0f 200. The observed frequency is 91, 75, 18, and 16. Determine the chi-square value and refer to a chi-square table. If your calculated value is greater than the table value (at .05 level) then the responses are not "similar" to what market analysts believe. If the calculated value is not significant (less than table value) then the observed frequencies are similar to the expected frequencies.

Re: determining if groups are similar to what's expected
Message posted by JG on November 19, 1999 at 12:00 AM (ET)

Your original model is too 'neat' and has probably been simplified to make it easier to remember and communicate to others. 50,40,5,5 is just too 'convenient' .

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.