Chapter 16 ppt eval & testing 4e formatted 01.10 kg edits
- 1. © 2013 Springer Publishing Company, LLC. Chapter 16 Interpreting Test Scores &Oermann Gaberson Evaluation and Testing in Nursing Education 4th edition
- 2. © 2013 Springer Publishing Company, LLC. Interpreting Test Scores ♦ A test produces a score – Number with no intrinsic meaning – Must be compared with something that has meaning ♦ Interpretations can be norm- or criterion- referenced 2
- 3. © 2013 Springer Publishing Company, LLC. Test Score Distributions ♦ Scoring a test produces a collection of raw scores, recorded by student name or number – Difficult to interpret characteristics of the scores ♦ Arrange in rank order, highest to lowest – Reveals range of scores – Still difficult to judge how a typical student performed on the test or other characteristics of the obtained scores 3
- 4. © 2013 Springer Publishing Company, LLC. Test Score Distributions ♦ Frequency distribution – Remove student names or numbers – List each score once – Tally number of times each score occurs – Identify how well the group of students performed on the exam more easily – Can represent graphically as a histogram or frequency polygon • Display scores that occurred most frequently, score distribution shape, range 4
- 5. © 2013 Springer Publishing Company, LLC. Characteristics of Score Distributions ♦ Symmetry ♦ Skewness ♦ Modality ♦ Kurtosis 5
- 6. © 2013 Springer Publishing Company, LLC. Symmetry ♦ Symmetric distribution or curve – Equal halves, mirror images of each other ♦ Nonsymmetric or asymmetric distribution or curve – Scores cluster at one end, tail toward other end – Most nursing test score distributions 6
- 7. © 2013 Springer Publishing Company, LLC. Skewness ♦ Skew—direction in which the tail extends – Positive skew—tail toward the right (in the direction of positive numbers on a scale) • Positively skewed distribution—cluster of scores at low end – Negative skew—tail toward the left (in the direction of negative numbers) • Cluster of scores at the high end • Most nursing test score distributions 7
- 8. © 2013 Springer Publishing Company, LLC. Modality ♦ Number of peaks (cluster of scores) in the distribution ♦ Mode – Most frequently occurring score in the distribution ♦ Unimodal—one peak ♦ Bimodal—two peaks ♦ Multimodal—many peaks 8
- 9. © 2013 Springer Publishing Company, LLC. Kurtosis ♦ Relative flatness or peakedness of the curve ♦ Platykurtic—relatively flat, gently curved ♦ Mesokurtic—moderately curved ♦ Leptokurtic—sharply peaked 9
- 10. © 2013 Springer Publishing Company, LLC. “Curving” Grades ♦ Not appropriate if scores lack characteristics of a normal curve – Bell-shaped: symmetric, unimodal, mesokurtic
- 11. © 2013 Springer Publishing Company, LLC. “Curving” Grades ♦ Most score distributions from teacher-made tests not normally distributed ♦ Shape of distribution affected by: – Test characteristics • Difficult test → positively skewed curve – Ability of students • Nursing content knowledge not normally distributed – Students admitted to nursing program not representative of general population 11
- 12. © 2013 Springer Publishing Company, LLC. Measures of Central Tendency ♦ Ways of indicating the score that is most characteristic or typical of the distribution ♦ “Middle” of a distribution, scores tend to cluster around it ♦ Three measures – Mode – Median – Mean 12
- 13. © 2013 Springer Publishing Company, LLC. Mode ♦ Most frequently occurring score in a distribution ♦ Must be an actual obtained score ♦ Identified from frequency distribution or graphic display without mathematical calculation ♦ Rough indication of central tendency ♦ Least stable measure of central tendency – Can fluctuate considerably among samples drawn from the same population 13
- 14. © 2013 Springer Publishing Company, LLC. Median ♦ Point that divides a score distribution into equal halves ♦ 50th percentile—50% of scores are above and 50% are below ♦ Does not have to be an actual obtained score – Even number of scores—median is halfway between the two middle scores – Odd number of scores—median is the middle score ♦ Index of location—not influenced by the value of each score – Good for skewed distribution 14
- 15. © 2013 Springer Publishing Company, LLC. Mean ♦ Mathematical average of all scores – Computed by summing individual scores and dividing by the total number of scores – Does not have to be an actual obtained score ♦ Value of the mean is affected by every score in the distribution – Influenced by extremely high or low scores – Not the most accurate measure of central tendency in highly skewed distributions 15
- 16. © 2013 Springer Publishing Company, LLC. Selecting a Measure of Central Tendency ♦ Relationship between shape of a distribution and locations of measures of central tendency – Normal distribution • Mean, median, and mode have the same value – Positively skewed distribution • Mean is highest, mode is lowest – Negatively skewed distribution • Mode is highest, mean is lowest 16
- 17. © 2013 Springer Publishing Company, LLC. Measures of Variability ♦ Used to determine how similar or different the test scores are ♦ Score distributions may have similar measures of central tendency and different degrees of variability ♦ Most common measures – Range – Standard deviation 17
- 18. © 2013 Springer Publishing Company, LLC. Range ♦ Simplest measure of variability ♦ Difference between the highest and lowest scores in the distribution – Sometimes expressed as highest and lowest scores, rather than a difference score (e.g., 42 to 60) ♦ Can be highly unstable—based on only two values ♦ Tends to increase with number of scores – Wider range of test scores from large group of students because of likelihood of an extreme score 18
- 19. © 2013 Springer Publishing Company, LLC. Standard Deviation (SD) ♦ Most common and useful measure of variability ♦ Takes every score in the distribution into consideration ♦ Based on differences between each score and the mean ♦ Represents average amount by which scores differ from the mean – Smaller if scores cluster tightly around the mean – Larger if scores widely scattered over large range 19
- 20. © 2013 Springer Publishing Company, LLC. Interpreting an Individual Score ♦ Scores on teacher-made tests – Norm-referenced interpretations • Use mean and SD to interpret individual scores – Criterion-referenced interpretations • Used in most nursing education settings • Scores are compared to a preset standard • Example: percentage-correct score – Comparison of a student’s score with the maximum possible score 20
- 21. © 2013 Springer Publishing Company, LLC. Percentage-Correct Scores ♦ Derived (not raw) score ♦ Often used as a basis for assigning grades ♦ Determined more by test item difficulty than by quality of performance – If test is more difficult than expected, teacher may want to adjust the raw scores before calculating the percentage correct ♦ Not to be confused with percentile score – Norm-referenced interpretation 21
- 22. © 2013 Springer Publishing Company, LLC. Interpreting an Individual Score ♦ Scores on standardized tests – Usually used to make norm-referenced interpretations – More relevant to general rather than specific instructional goals • Should not be used to determine course grades – Usually reported in derived scores • Percentile ranks • Standard scores • Norm-group scores (cont’d) 22
- 23. © 2013 Springer Publishing Company, LLC. Interpreting an Individual Score ♦ Scores on standardized tests (cont’d) – Important to specify an appropriate norm group for comparison – User’s manual includes norm tables with descriptions of each norm group – Teacher should select the norm group that most closely matches the group of students • Examples: type of nursing program, public or private 23