These are potentially stupid questions, but bear with me – I am not a statistician nor a mathematician.

If a group of observations is normally distributed, will the mean value of that group also follow a normal distribution? Furthermore, if the mean values of two groups are both normally distributed, will the difference between these mean values also follow a normal distribution?

From my understanding, the student's t-test does not actually assume that observations in the two groups are derived from normally distributed populations, but rather that the difference between the mean values of these groups are normally distributed (or at least approximately). Hence, if the answer to both my questions is ‘yes’, then it makes sense to me, why one would still apply different methods, such as histograms and the K-S test, to assess whether the data follow a normal distribution (especially in studies with small sample sizes, where the central limit theorem is less likely to ‘kick in’). ]]>

I have built a logistic regression model with misclassification rate around 18%. When I look at the ROC curve(area under the curve is 0.76) and find out the accuracy at various cutoff points, the best accuracy is at 0.3 probability cutoff. Is 0.3 acceptable?Shouldn't the cutoff chosen be greater than 0.5?

Thanks ]]>

I'm a freshman here, but i need a help.

I need to run a ANOVA + Post Hoc test using filters by "select cases" tool but it's not working

I already used the 6 generated filters trying an ANOVA as fixed factors and with one dependent variabble "Price", but the errors are:

-the post hoc dosen't ran saying there are few than two groups

- the complementar filter (filter 1 = China Low filter 2 = China High) are showing non complementar means, because filter 1 shows Not Selected Mean = 4,5 and the filter 2 Selected Mean shows a different value o mena. They should be the same, right?

Summarizing, what I need to do to run the ANOVA and Post-Hoc with those 6 filters and with one dependent variabel at the same time?

Thanks! ]]>