Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Significant test please help

  1. Aug 12, 2012 #1
    Significant test!! please help

    I have two questions.

    1) if two sets of data have different sample numbers, then can we perform any significant test (e.g., anova, tukey had) to see whether there is a significant difference between the two sets of data?

    2) This is a totally different question with (1). 80 blind people picked out one of 4 fruits from a basket. The results show that 64 blinds picked an apple, 16 blinds picked a strawberry, 13 blinds picked a banana, and 7 blinds picked an orange. Can we discover if the apples were picked by a chance or it has some statistical meaning? If so, can we use Chi square test?

    Thanks in advance.
  2. jcsd
  3. Aug 15, 2012 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Re: Significant test!! please help

    1) It's not clear what you are asking. Is it this: the two sets of data are independent samples from two different populations, and you're trying to test whether the populations have the same distribution?

    2) Passing over the fact that a blind person would have no more difficulty telling the fruit apart than would a sighted person:
    - your stats add up to 100, yet each only took one?
    - were there equal numbers of each fruit? Or is the point of the exercise to tell if there were different numbers?
    - you can never use statistics to say whether something happened by chance; what you can do is estimate the probability of getting a specified result by chance from a specified distribution.
  4. Aug 15, 2012 #3


    User Avatar
    Science Advisor

    Re: Significant test!! please help


    There are significance tests for difference of means by using sample data. Is that what you're looking for?

    Just a small nitpick that , in 2), 80≠ 64+16+13+7=100 . Did some people choose more

    than one fruit?

    If you want to test whether the frequency of a fruit chosen is statistically the same

    as the frequency (distribution-wise ) of the fruits in the basket, you can do that, but

    you need to have the distribution of the fruits in the basket and then run your test,

    and decide if the choice of fruits was or not random.
    Last edited: Aug 15, 2012
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook