Chapter 9: Misconceptions
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 9.
- When testing a single correlation, the null hypothesis
must be set up to say that the correlation in the population is equal
to 0.00.
- If a researcher asserts that "p < .001"
when testing Ho: r = 0, you can be confident that the sample value of
r was closer to +1.00 (or -1.00) than 0.
- The sampling distribution of r is symmetrical.
- If a researcher sets alpha equal to .05 when testing
each of the correlation coefficients in a correlation matrix, he/she
has a 1-in-20 chance of making a Type I error.
- It's important for researchers to test reliability
and validity coefficients against a null hypothesis that says Ho: r
= 0.
- If a correlation coefficient turns out to be statistically
significant, there's no need to compute r2.
- Since the "error" associated with less-than-perfect
measuring instruments is considered to be random (rather than systematic),
this kind of error will "balance itself out" and not cause
correlation coefficients to be systematically higher or lower than they
ought to be.
- If the test of a correlation coefficient turns out
to be statistically significant, this indicates that the mean on variable
X is significantly different from the mean on variable Y.
- Tests of Pearson's r are robust to the underlying
assumptions of linearity and homoscedasticity.
|
