Common Flaws in Multiple-Choice Tests

From Clinicaltools.com

Jump to: navigation, search

Each course objective should have one knowledge question. This applies to regular post-tests and evaluative pre/post-tests. Remember to write questions that people probably would answer incorrectly, if they didn't take your course.

  1. Weak Stem: the stem should present a clearly-defined problem, usually specific enough to stand alone as a completion item.
  2. Heterogeneity of Options: the choices should all be of similar nature; otherwise, one or more will be implausible.
  3. Grammatical Cues: all options should be grammatically consistent with the stem.
  4. Lack of Parallelism: the logical flow of both grammatical construction and ideas should be consistent from stem to option.
  5. Complicated Stem: these should be avoided since stems which include several ideas can be open to different interpretations by different examinees.
  6. Complicated Options: these should be avoided since it is usually too difficult a task to make rapid comparisons among complex options.
  7. Explanatory Stems: stems which "teach" material not necessary to the question should not be included, as examinees are supposed to learn from the content of the course and not the test.
  8. Negatively Stated Stem: stems usually present the task of choosing the "correct" or "best" option; a change in ground rules should be avoided and, if absolutely necessary, the change should be emphasized by using bold, underline, color, etc.
  9. Repetition of Option Lead-ins: if all options begin with the same wording, it should be included at the end of the stem.
  10. Specific Determiner: words or phrases which tend to be present in only true or only false statements (i.e., always, never, etc.) should be avoided in the options.
  11. Choices Not at the End of the Stem: items should be written so that the natural location of the answer is at the end of the stem.
  12. More Than One Acceptable Answer: only one option should be best; the others should be clearly unacceptable.
  13. Use of "All of the Above" as an Option: this option would likely be selected when at least two of the other choices are correct, allowing an examinee to receive credit for partial knowledge.
  14. Clues from Other Items: no item should be answerable through use of information contained in other items.
  15. Pattern of Keyed Answers: the position of the keyed option should not follow any discernible rule throughout the test; the keyed position should be determined randomly or by a consistent practice such as placing all options in alphabetical or numerical order for each item.
  16. Oner-inclusive Options: if one or more options cover all possibilities, the choice will likely be made at the expense of other options.
  17. Inconsistent Lengths of Options: usually the longer option tends to be the correct one; try to make all options of similar or equal length.

Examples of Good Multiple-Choice Questions

Here are a few examples of well-written test items taken from one of the courses on GeneticSolutions.com. These serve as good models:

1. In cancer cells, oncogenes function to:
___ slow cellular growth
___ repair DNA more quickly
___ promote more rapid cellular growth
___ generate greater amounts of cellular energy
___ create greater amounts of RNA
2. In healthy cells, proto-oncogenes function to:
___ slow cellular growth
___ repair DNA
___ promote cellular growth
___ generate cellular energy
___ create RNA
3. Sporadic cancers tend to:
___ occur at younger ages than do hereditary cancers
___ metastasize more easily than do hereditary cancers
___ occur more frequently in Ashkenazi Jewish individuals
___ be unilateral and not bilateral
4. Hereditary cancers tend to:
___ be inherited in an autosomal recessive pattern
___ occur at younger ages than do sporadic cancers
___ be passed more frequently from mother to daughter than from father
___ be unilateral and not bilateral