Popham Ch. 13 & Ch. 14 Pondertime

Chapter 13

1.) If you had to use only one of the three individual student interpretation schemes treated in the chapter (percentiles, grade-equivalent scores, and scale scores), which one would it be? Why did you make that choice?

That would depend on my objectives or rather the kind of inferences I needed to make:

Percentiles correlate a student’s score to a norm group. If I had evidence of predicted performance of the norm group, let’s say for a program I had, I could use the percentile rank to determine eligibility for an advanced placement program.  For example, if every person that ranked below the 75th percentile failed the AP program, I could structure admittance into the AP program around such a standardized test score.

Grade equivalent scores are a little more limiting in their use. I can see using a grade equivalent to help determine a baseline from which to work from. For example, if a student is scored below their grade level in reading comprehension, I can differentiate content and tests consistent with that grade level for those students. Grade level rankings at or above a student’s actual grade level is really useless information.  This is because a score of 6.3 on a reading comprehension test for example, for a fourth grader tells me that a sixth grader in her third month of the term would score similarly, but does not tell me the fourth grader is comprehending at a sixth grade level.

Scaled scores would be useful if I needed to make inferences comparing students that took equivalent, but different tests. Since scaled scores normalize scores across different yet similar tests. This type of scoring would be useful for longitudinal tracking.

2.) It is sometimes argued that testing companies and state department of education adopt scale-score reporting methods simply to make it more difficult for everyday citizens to understand how well students perform. (It is easier for citizens to make sense out of “65% correct” than “an IRT scale score of 420”). Do you think there is any truth in that criticism?

Yes, it is easier for people to make sense out of a “65% correct” than an IRT scale score of 450, if people are not informed about the nature of an IRT scale score. If understood an IRT scale score is more practical for a parent than the “65% correct score. This is because as mentioned above, scaled scores are normalized raw scores. If a student scored 65% on tests from one year to the next that does not mean they performed the same from one year to the next. This is because the two tests taken were different and even though the student received, the same score does not necessarily mean the student performed the same. Yet, a scaled score of 450 is a normalized across tests. The score is normed to a standard and thus means the same scores are an equivalent performance from test to test.

Chapter 14

1.) Can you think of guidelines, other than the two described in the chapter, to be used in evaluating a classroom teacher’s test-preparation practices? If so, what are they?

Test preparation should be content driven.

2.) How about test-preparation practices? Can you think of any other sorts of test-preparation guidelines that are meaningfully different from the five described in the chapter? If so, using the chapter’s two evaluative guidelines or any new ones you might prefer, how appropriate are such test-preparation practices?

No.

Popham, J.W. (2011). Classroom Assessment. What teachers need to Know. Sixth Edition. Pearson. Boston.

This entry was posted in L1 – Learner centered, P1 – Informed by professional responsibilities and policies, P2 – Enhanced by a reflective, collaborative, professional growth-centered practice, S1 – Content driven, S2 – Aligned with curriculum standards and outcomes, T2 – Intentionally planned, T3 – Influenced by multiple instructional strategies and tagged , , , , . Bookmark the permalink.

3 Responses to Popham Ch. 13 & Ch. 14 Pondertime

  1. I love your replies; you provide as much information as the questions warrant. I continually make the mistake of trying to answer simplistic or leading questions with deeper thoughts when in fact the only result with “simple” questions is to waste time.
    Thanks!

  2. merrylpohl says:

    David,
    I agree that a percentage score is easier to read and translates well to most people. I see your point of how scale scores make it easier to compare scores across different tests. I feel that it can be difficult to understand the meaning of a scaled score and that is what makes them confusing for most people to read.
    -Merryl

  3. Jenny says:

    David,
    Thanks for your clear explanations of grade level scores and IRT scaled scores. Most test results come with guides that show how to interpret scores, so I would hope that parents would consult that or ask a teacher to explain their children’s scores.

Leave a comment