Identify the threats to internal validity illustrated in the examples below. Students falling below ...

Identify the threats to internal validity illustrated in the examples below. Students falling below the 10th percentile on a reading achievement test later receive higher scores on a similar test; students above the 90th percentile receive lower scores on the second test administration.
 
  a. experimental mortality
  b. reactive effect of testing
  c. statistical regression
  d. maturation
  e. instrumentation

Ques. 2

The experimental approach to research starts with groups that are ____ on significant variables, while the ex post facto approach starts with groups that are ____ on significant variables.
 
  a. different; equivalent b. equivalent; different

Ques. 3

Which of the following would represent a variable that could be manipulated in an experiment?
 
  a. Lecture vs. discussion method of teaching
  b. High vs. low socioeconomic status of subjects
  c. Male vs. female subjects
  d. College prep vs. vocational curriculum
  e. High vs. low scorers on an aptitude test

Ques. 4

Which of the following variables can be either an active or an attribute variable in a research study?
 
  a. verbal aptitude c. creativity
  b. anxiety d. reading readiness

Ques. 5

Matching of experimental and control subjects, homogeneous selection, and analysis of covariance can all be used to
 
  a. reduce the effects of statistical regression.
  b. reduce initial differences between treatment groups.
  c. counterbalance the effects of experimental treatments.
  d. control the learning to learn effect.

Ques. 6

The most important reason for choosing an ex post facto design rather than an experimental design is when
 
  a. random sampling is not possible.
  b. random assignment must be used.
  c. manipulation of the independent variable is not possible.
  d. intact classroom groups must be used.
  e. young children are the subjects in the research.

Ques. 7

Matching subjects in the experimental and control groups is useful when
 
  a. small samples are to be used.
  b. large differences between the groups on the dependent variable are likely.
  c. random assignment to groups is not possible.
  d. All of these are true.
  e. a and c

Ques. 8

High school graduates have a higher average income than those who did not graduate from high school. In this situation we can rule out:
 
  a. common cause. c. Both of these answers.
  b. reverse causality. d. Neither of these answers

Ques. 9

In experimental research, a random procedure is recommended for the
 
  a. selection of subjects. c. ordering of multiple treatments.
  b. assignment of subjects to treatments. d. Al of these are true.