Properties of Souvenirs at two Destinations

timtak

New Member
#1
Dear Statisticians

To discover if perceived preferences in souvenirs are different, we collected 31 souvenirs from Guam and 30 from Tokyo, on the net and counted how many:

Depict symbols associated with the place (such as coconut palms / Tokyo Tower)
Mention or depict the deliciousness of the souvenir
Use a mascot
Mention the place name (Guam/Tokyo)

The results were
Guam: 19 25 3 22
Tokyo: 10 12 16 20

Please will someone be so kind as to tell me, whether it was appropriate of me to use four 2x2 Fishers Exact Tests to calcuated whether Guam and Tokyo souvenirs were statistically different in each of the four properties, or whether I should have used some other test?

Incidentally the results were
p=0.401 p< .05 *
p=0.0016 p< .05 *
p=0.0003 p< .01 **
p=0.7963 n.s.

I used an online 2 x 2 Fishers Exact Test tool but I seem unable to paste the URL probably because I should not.

Thank you to forum members for your kind help in the past,

Tim
Timothy Takemoto
Yamaguchi University
 

Karabiner

TS Contributor
#2
Yes, you can use the 4 Fisher tests. I guess the first result was 0.0401, not 0.401?
One could think about doing a correction for mutliple testing, e.g. Bonferroni
(new alpha would then be 5%/4=0.0125).

With kind regards

Karabiner