Levene's Test for Equality of Variance

[maccaman]

I have just started University and we have been doing t-tests however in our Computer labs we are supposed to do the Levenes Test. I have done the levenes test from our data of males and female heights in our class. i was just wondering what a value of F = 0.001 and Sig. 0.982 would mean. the computer program we are using is SPSS. I know its about testing if the variances are the same, however i don't know what the values are telling me. Any help would be greatly appreciated.

 

[xanthym]

I have just started University and we have been doing t-tests however in our Computer labs we are supposed to do the Levenes Test. I have done the levenes test from our data of males and female heights in our class. i was just wondering what a value of F = 0.001 and Sig. 0.982 would mean. the computer program we are using is SPSS. I know its about testing if the variances are the same, however i don't know what the values are telling me. Any help would be greatly appreciated.

Statistical tests usually involve making various assumptions about the characteristics and parameters of the population being sampled. In your case, you are performing the "t-test" to test hypotheses about 2 sampled groups. One of the assumptions made by the standard "t-test" is that the 2 populations being sampled have EQUAL VARIANCES. The purpose of the Levenes Test is to test and verify that this equal variance assumption is reasonable.

The Levenes Test outputs 2 parameters in SPSS. The first is the F statistic value. The larger the F statistic number, the greater is the possibility the variances are different. Similarly, the smaller the F value, the greater is the probability that the variances are equal. In your case, F appears very small (which is good ).

The probability that the variances are equal are reported in the value labeled "Sig", which stands for "significance". This number is a probability between 0 and 1, and the closer it is to 1, the greater is the probability the variances are equal. In your case, the "Sig" value (0.982) is very high and indicates a high probability the population variances are equal. Knowing this, you can now use the standard parametric "t-test" (which makes the assumption of equal variances) with confidence that the variances are most likely equal.


~~

 

[Umara00]

Hi
Im a lil confused here.. the thing is I am getting no F value in my SPSS output. Like the the F value area is blank. What do i do?