Turn on thread page Beta
    • Thread Starter
    Online

    9
    ReputationRep:
    Hi,

    Can someone check this? I have attached the SPSS output regarding the Levene's test and ANOVA test.

    According to this, the Levene's test is significant meaning equal variances was violated and homogeneity of variance cannot be assumed, F(3,60) = 2.99, p < .05. Is this correct?

    Main effect 1 of humourosity on perceived stress:
    F(1, 60) = 12.19, p <.001.

    Main effect 2 of gender on perceived stress:
    F(1, 60) = 5.15, p <.05

    Interaction effect between humourosity and gender on perceived stress:
    F(1, 60) = .048, p >.05.

    Thanks.
    Attached Images
      
    Offline

    18
    ReputationRep:
    (Original post by DianaJones)
    Hi,

    Can someone check this? I have attached the SPSS output regarding the Levene's test and ANOVA test.

    According to this, the Levene's test is significant meaning equal variances was violated and homogeneity of variance cannot be assumed, F(3,60) = 2.99, p < .05. Is this correct?

    Main effect 1 of humourosity on perceived stress:
    F(1, 60) = 12.19, p <.001.

    Main effect 2 of gender on perceived stress:
    F(1, 60) = 5.15, p <.05

    Interaction effect between humourosity and gender on perceived stress:
    F(1, 60) = .048, p >.05.

    Thanks.
    It's better to report p values exactly (e.g. Interaction effect between humourosity and gender on perceived stress:
    F(1, 60) = .048, p = .83.) F & p should also be italicised.

    The issue of unequal variances is a bit tricky (see andy field's textbook 3rd edition p.360). The fact that levene's test is significant raises concerns that the rest of your results are invalid. According to Andy Field if the sizes of the groups are roughly equal then the ANOVA test is fairly robust to having unequal variances.
    • Thread Starter
    Online

    9
    ReputationRep:
    (Original post by iammichealjackson)
    It's better to report p values exactly (e.g. Interaction effect between humourosity and gender on perceived stress:
    F(1, 60) = .048, p = .83.) F & p should also be italicised.

    The issue of unequal variances is a bit tricky (see andy field's textbook 3rd edition p.360). The fact that levene's test is significant raises concerns that the rest of your results are invalid. According to Andy Field if the sizes of the groups are roughly equal then the ANOVA test is fairly robust to having unequal variances.
    Thank you, that helps a lot.

    It's strange, because we were told to obtain an equal number of participants in each condition. Yet, this is what is causing it to have unequal variances which is definitely not a good thing. I don't understand why it does this.
    Offline

    18
    ReputationRep:
    (Original post by DianaJones)
    Thank you, that helps a lot.

    It's strange, because we were told to obtain an equal number of participants in each condition. Yet, this is what is causing it to have unequal variances which is definitely not a good thing. I don't understand why it does this.
    No having unequal variances isn't an effect of having equal group sizes... if your sample sizes are more equal then the unequal variance assumption matters less since the tests are more robust.
    • Thread Starter
    Online

    9
    ReputationRep:
    (Original post by iammichealjackson)
    No having unequal variances isn't an effect of having equal group sizes... if your sample sizes are more equal then the unequal variance assumption matters less since the tests are more robust.
    What do you mean by the tests being more robust? So, that means that the results are not likely to be invalid?

    Posted from TSR Mobile
    Offline

    18
    ReputationRep:
    (Original post by DianaJones)
    What do you mean by the tests being more robust? So, that means that the results are not likely to be invalid?

    Posted from TSR Mobile
    Yeah but you didn't provide this infomation in your original post. Also what matters is the degree to which the variances are unequal which you havent really told us (this is the ratio of the smallest variance (SD^2) to the largest variance).
 
 
 
Poll
Do you think parents should charge rent?

The Student Room, Get Revising and Marked by Teachers are trading names of The Student Room Group Ltd.

Register Number: 04666380 (England and Wales), VAT No. 806 8067 22 Registered Office: International House, Queens Road, Brighton, BN1 3XE

Write a reply...
Reply
Hide
Reputation gems: You get these gems as you gain rep from other members for making good contributions and giving helpful advice.