A list of eight thveats te internal validity is presented by the
authors, along with 4 threats to urternal validity. Seven of the
threats to internal validity are given a dotsilei discussion in this
section of the paper.

»History- the offect of events sxtemal to the experiment on subjects
| betužen measurements. History becomes a more plausible oonfamding
factor, the longer the tine betusen measuronente,
sHaturatione The nature of the subjoeto changes over time, Obviously,
this can cecur in most any netural system,
»fest offoctse the effects of one measuroaant on suooseding measurements,
The familjar example of improving GMAT secres on the second try wuld
fit here, A subset of ooncern here is that of resetivity, Since
neasurenent may change what, is being measured, one wuhnts non«roactive
ueasures; measures that will not cause a change in the subject,
sInstramentatlone changos in the measuring instrument between
meszuriants, This could include the scoring of measures, as in using
different judges to categorize open«emded responees on onestionnaires,
»Statistical Regression (aka regression towards the meen)e the
movanent of postetost measurements of oxtnome grospe tovards the pre-
test mean of all groups. (Regressicn tovards the mezn is a very
inportent eoncept in evolutionary biology)
»Selection- the differences betswen groupa nay be dus to recruitnent ra
rather than treatnent.
»Mortality- differences betwes gronps due to differential drop-out.
We discussed an example of this under subjoets reactions to experimentil
trsatnonte.