What is the relationship between external validity and the college sophomore problem? What will ...

External validity is the extent to which the results of an experiment can be generalized. Because most research in psychology is completed on college sophomores, generalizability (or external validity) can be problematic.

- - - - - - - - - - - -

In a proportionate stratified sample, each of the subgroups in the sample has exactly the same proportion as the corresponding subgroup in the population. For example, if 14 of the population is between 30 and 39 years old, then exactly 14 of the sample will be in this same age range. Thus, the composition of the sample exactly matches the composition of the population. A stratified random sample has equal representation from each of the specified subgroups. For example, if there are four subgroups in the population, then each will form 25 of the sample. However, the subgroups are probably not equally represented in the population. For example, one group may comprise only 10 of the population but will still make up 25 of the sample. Thus, the composition of the sample does not accurately represent the composition of the population. Finally, simple random sampling lets pure chance determine who is selected. As a result, a group that makes up 10 of the population can be severely overrepresented or underrepresented in the sample, depending entirely on chance. Thus, there is no guarantee that the composition of the sample will accurately represent the composition of the population.

- - - - - - - - - - - -

b

- - - - - - - - - - - -

A participant effect is a threat to internal validity in which the participant, consciously or unconsciously, affects the results of the study. A single- or double-blind experimental procedure can help to control for this. An experimenter effect is a threat to internal validity in which the experimenter, consciously or unconsciously affects the results of the study. Single and double-blind experimental procedures can help control for this also. A single-blind procedure controls for either participant or experimenter effects whereas a double-blind procedure controls for both at the same time.

- - - - - - - - - - - -

A proportionate stratified random sample is appropriate when researchers are interested in the entire population. This type of sampling ensures that the composition of the sample will provide an accurate representation of the composition of the population. Note, a proportionate stratified random sample is a probability sample, so the entire population must be known.

- - - - - - - - - - - -

a

- - - - - - - - - - - -

Instrumentation is a threat to internal validity in which changes in the dependent variable may be due to changes in the measuring device. Mortality, or attrition, is a threat to internal validity in which differential dropout rates may be observed in the experimental and control groups, leading to inequality between the groups.

- - - - - - - - - - - -

A stratified random sample is appropriate when researchers want to describe and compare different subgroups that make up a population. This type of sampling ensures that each subgroup will be represented with a guaranteed number of participants. Note, a stratified random sample is a probability sample, so the entire population must be known.

- - - - - - - - - - - -

d

- - - - - - - - - - - -

Testing is a threat to internal validity in which repeated testing leads to better or worse scores (practice or fatigue effects). Regression to the mean is a threat to internal validity in which extreme scores, upon retesting, tend to be less extreme, moving toward the mean.

TY