In creating learning outcomes and test items, it is
imperative that the test items are valid measurements of the learning outcomes.
This may seem like common sense, but invalid measurements occur more often than
one might think, and it can be easy to mistakenly create tests that do not
truly measure what they are intended to do. For instance, as I stated in the
previous post, my learning outcomes are not intended to measure a student’s
reading ability. Although reading ability is highly important, my purposes of
the learning outcomes learn toward comprehension of literature. I’m not looking
for reading ability; I am looking for understanding of literature components,
such as the parts of a book, the main idea, and character differentiation. At this stage, I want to make sure my
students are developing an understanding of these components and what is read,
otherwise words are simply words with no real meaning. Because this is my
intent, all written tests devised under these learning outcomes shall be read
to the students. If students are left to read the tests independently, there is
a chance of inaccuracy of test results because students that have mastered the
content but not yet mastered the skill of reading would be clearly at a
disadvantage. This would reflect negatively on test scores and would
inaccurately portray that they have failed to master the content.
With that in consideration, let’s examine the different test
items. For learning outcomes one and two, my test items were based on multiple
choice and short answer responses. The multiple choice questions have been
examined for grammatical cues and unintentional trickery. My goal is to measure
student’s ability to identify the parts of the book, not to try to trick them. The
pictures provided are intended to provide realistic application. Although this is
not a performance assessment, I want to give students the most realistic
examples in this written assessment. If possible, all children would be
provided a copy of the book and answer the questions by examination of the
book, but in reality picture cues are easier to implement. The disadvantage to
such format is that children are limited to select pages that are pictured, so the
answer choices are limited, but it does allow for measurement of identification
of the parts of a book. Short answer responses are also added to gain more
insight of learning outcomes. With multiple choice questions there is always a possibility
that students will simply guess at the answer. In order to provide a more
accurate view of student understanding, short answer questions are added so
that students must provide an answer. The questions do not require higher-level
thinking in these assessments because the intent is to measure at the knowledge
level in identifying the components.
Creating an essay item presented more of a challenge for me.
Although there are great benefits in essay items, devising an item for second
grade students seemed a little difficult. Typically at this stage, students are
beginning to understand the purposes of writing and essay formats are not
generally used. It’s not that it is not possible for second grade student to
write an essay; it’s simply that they are still in the learning how to write
stage. For this reason, it is important to acknowledge that there will be
spelling and grammatical errors. Most children in this grade level understand
that there must be capital letters at the beginning of a sentence and
punctuation at the end, but they are still learning how to spell, how to
construct sentences with a subject and predicate, how to capitalize proper
nouns, and so forth. Obviously they will not be free from errors. For this
reason, although instilling grammatical rules and proper spelling are
important, the use of the essay in this assessment is to measure the ability to
understand character traits and comparison among characters. In such an
assessment, points shall not be lost due to such errors and should focus on the
content of the essay rather than the structure. If the essay was intended to
measure grammatical skills in writing, then by all means errors should be
included in scoring, but for the intent of this essay, it would not be valid to
do so.
For the essay item of choosing two characters to compare
similarities and differences, higher-level thinking skills are used to analyze
and compare and contrast the characters. They essay will reflect student’s
understanding of character traits and how characters are alike and different,
thus reinforcing comprehension skills. The choice of text does not involve a
wide array of characters, so students will not have to divulge too much time in
choosing which characters to use. Rather than pick the characters for the
students, I prefer to allow students to make the choice. The scoring will not
be based on which character they choose; rather scoring will be based on the
ability to describe the similarities and differences among the two characters
they chose. Scoring would be rubric-based with a rating scale based on at the
essay consisting of at least three characteristics described for similarities,
at least three characteristics described for differences, the descriptions
accurately portraying the character of their choosing, and the analysis of the
characters.
The testing measurements I chose are intended to measure the
learning outcomes. I chose these methods because I feel that they have more
advantages in assessing what I want to measure. Although I considered using
matching in identifying book parts, I felt that at this stage there would be
too many options and I did not want to overwhelm students with too many
choices. Multiple choice assessments provide a smaller sample of choices that
allow for measurement of knowledge. On a similar note, I chose to avoid true/false
questions because I want to ensure that students are identifying the correct
book parts. True/false questions are more susceptible to guessing and I did not
feel would be advantageous in deciphering if students have mastered identification
of the content.
As for the essay response, because I intend to measure the
higher level thinking skill of analyzing the characters, this method seemed to
be the best measurement tool. Kubiszyn and Borich (2010) point out that
although essays are time consuming and can be difficult to score, the essay is
highly beneficial for measuring high-level cognitive skills. In using the
essay, the answers are not there for students to quickly identify or pick from
a set; students must demonstrate their own knowledge of the topic. For this
reason, I felt it would be difficult to utilize most other forms of written
assessment to measure analysis of characters.
References:
Kubiszyn, T., Borich, G. (2010). Educational testing & measurement (9th
ed). Hoboken, NJ: John Wiley & Sons, Inc.