A. Constructed response questions:
are open ended, short answer questions that measure
application-level cognitive skills as well as content knowledge. No opinion questions.
use a range of primary and secondary stimuli and authentic
"real world" examples including time lines, maps, graphs, cartoons, charts, and
short readings.
are graded against specific criterion (employ a scoring rubric).
B. Constructed response questions should be based on a given
stimulus and include a series of questions which build from simple to complex.
First question: A general or specific question with the answer found
in the stimulus. "What was the population in the year?"
Second question: Make connections between and among the
different parts of the stimulus. "In which year was the immigration the
greatest?"
Third question: Require the student to respond with information
related to (but not included in) the stimulus. "State a historic trend that
explains the data?"
C. Constructed response questions can assess higher level thinking:
- comparisons, contrasts, causes, effects, changes
- identify patterns or conflicting points of view
- categorize or summarize information
- construct graphs or charts from data
- state a generalization, conclusion, explanation or prediction
D. Scoring: Each constructed-response question should be scored
according to a rubric (scoring guide) that gives varying degrees of credit
for correct or partially correct answers. The rubric should include enough information or
examples to allow different raters to arrive the same score for a given student response.
1. A rubric for a short constructed-response question can be:
"Complete" score (2) represented a complete and
appropriate answer.
"Partial" score (1) indicated that the response had some,
but not all, of the components of an appropriate response.
"Inappropriate" score (0) represented an answer that had
none of the components of an appropriate response.
2. Extended constructed-response questions are lengthier and more
complex exercises that allow for a finer level of discrimination in scoring the responses.
Their rubric can be:
"Complete" score (3) was assigned to a response that was
complete and appropriate.
"Essential" response (2) was less complete but included
the most important components of an appropriate response.
"Partial" response (1) included some appropriate
components, but fewer or less central ones than those required for an
"Essential" score.
"Inappropriate" (0) response included only inappropriate
material.
|