When people read, hear, or prepare research summaries,
they sometimes have misconceptions about what is or isn't "sound
practice" regarding the collection, analysis, and interpretation
of data. Here are some of these common (and dangerous) misconceptions
associated with the content of Chapter 10.
In comparing the means of two groups, the null hypothesis
must be set up to say that one population mean is equal to the other
If the means of two groups are statistically compared
and found to be significantly different, those two means must be quite
A researcher who compares two means with an F-test
is more sophisticated than the researcher who compares two means with
Correlated t-tests focuses on correlation coefficients.
Two-tailed t-tests are used when 2 groups are compared
whereas one-tailed tests are used in conjunction with one-sample t-tests.
If two groups are compared with an analysis of variance,
the researcher's primary interest is in the comparative degree of variability
in each group.
Two means that are significantly different at p <
.001 must be further apart than two means that are significantly different
at only p < .05.
A test that has adequate power to detect large effects
has even greater power to detect small effects.
When interested in comparing the means of two samples
that differ in size, smart researchers discard data from the larger
group in order to make the sample sizes equal which then causes the
t- or F-test to be robust.