Pre-test/Post-test Guidelines

Written by Dr. Rolfe Windward

One of the ways to provide evidence of teaching impact is to (a) give students a pre-test to evaluate their current level of knowledge about a subject, (b) followed by instruction on that subject and (c) ending with a post-test to evaluate degree of learning. There are many ways to do this but if you want to make the results comparable across subjects or classes you need to follow some basic rules:

- The pre- and post-test for a given subject must contain essentially the same questions although these can be altered or masked in various ways and the same scoring guide should be used to evaluate both tests.
- There
must be a way to calculate a percentage change in performance between the
pre-test and post-test. This naturally assumes a numerical scoring method
associated with the scoring guide (rubric); that is, however you evaluate
the tests, you need to be able to express the results as a number.
**This does not mean you should only use 'objective' tests**(multiple choice, etc.) - objective tests are actually not easy to design well and narrative descriptors in a rubric may provide a better indication of the level of student competence in many cases; you just have to be able to rate student performance in a manner that can be translated into numbers. - Those
numbers can then be averaged and divided by the total to show a percentage
change and that is what you are looking for - the degree of improvement in
student knowledge expressed as a percentage. You should do that for each
student and also for the class as a whole (see
__Recording the Results__section below).

__Recording the Results__

Tabulate the results for the entire class (raw scores are in parenthesis in the example below):

Teacher Name: Content Area:

Student Name |
Pre-test Score% (5) |
Post-test Score% (5) |
Difference |

Student A |
.4 (2) |
.6 (3) |
.2 (1) |

Student B |
.6 (3) |
.8 (4) |
.2 (1) |

Student C |
.4 (2) |
.8 (4) |
.4 (2) |

Total Mean Change |
.466 (2.33/5) |
.733 (3.66/5) |
.267 (1.33/5) |

In the above example, both the pre-test and post-test have the same number of key core content questions (note that is not necessarily the same as the total number of questions in each test, you can vary that to mask the questions you wish to focus on). The last column is just subtracting the pre-test result from the post-test result.

Note: There are several different methods you can use to tally the results but don't mix them up. On the pre-test results in the example above you could just total up the student pre-test scores for the three students, 2+3+2=7, then divide by the total points in all three tests, 7/15=.466. Or you could average the student scores, 7/3=2.33 and then divide by the total number of key questions, 2.33/5=.466. Or you could add up the frequencies for the three students .4+.6+.4=1.4/3=.466. Same result regardless. Just be consistent and there will be no problems.