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 4.
If reliability is high, then validity most likely
is high as well.
If Researcher A uses a test that Researcher B built
a year ago and documented as being reliable, then Researcher A can use
that test today and have full confidence that it will be reliable for
High test-retest reliability implies high internal
consistency reliability, and vice versa.
If Larry's IQ score is 4 points higher than Mary's
IQ score, then we have a legitimate right to say that Larry is more
intelligent than Mary.
Test-retest reliability remains fairly stable regardless
of how much time passes between the test and the retest.
The "criterion" in criterion-related validity
is a numerical cut-off that is used to determine whether a measuring
instrument is or isn't valid.
Discriminant validity is documented best by correlation
coefficients that are near -1.00.
A study's findings should be considered to be valid
if that study's data are shown to have high reliability and validity.
The typical researcher spends as much time considering
how to collect data as he/she does considering how
to analyze data.
Kendall's coefficient of concordance (W) would turn
out equal to the mean of the various rs values if Spearman's rho is
computed for each pair of things being ranked.