Question 1

Possible applications of IRT were discussed in your textbook.Which of the following is NOT one of those possible applications?
A) determining measurement equivalence across testtaker populations
B) identifying a common metric among several tests measuring the same construct
C) evaluating existing tests for the purpose of mapping test revisions
D) developing item banks

Accepted Answer

The answer of Possible applications of IRT were discussed in...

Question 2

A math test developer is interested in deriving an index of the difficulty of the average item for his math test.As his consultant on test development,you advise him that this index could be obtained by:
A) identifying the item deemed to be average in difficulty and then deriving an item-difficulty index for that item.
B) averaging the item-difficulty indices for all test items and then dividing by the total number of items on the test.
C) dividing the total number of items on the test by the average item-difficulty index.
D) raising that very same question to a more knowledgeable test development consultant.

Accepted Answer

The correct approach to deriving an index of the difficulty of the average item for a math test is to average the item-difficulty indices for all test items and then divide by the total number of items on the test. This method ensures that the difficulty level of each item is considered, and the overall average difficulty is calculated accurately.

Question 3

With regard to item-discrimination indices,a d equal to -1 is
A) a test developer's dream.
B) a test developer's nightmare.
C) a testtaker's dream.
D) an insomniac's nightmare.

Accepted Answer

A discrimination index (d) of -1 means that every student who responded correctly to the overall test failed to respond correctly to the specific item in question. Therefore, the item is extremely easy and does not discriminate between high and low performers. This would be a nightmare for test developers, as the item would not provide any useful information about student ability.

Question 4

On the item characteristic curves for a test of ability,a large number of items biased in favor of male testtakers is found to coexist with the exact same number of items biased in favor of female testtakers.Based on these findings,it would be reasonable for the test developer to claim that the test
A) measures the same ability in the two groups.
B) is a fair test as any observed bias balances out.
C) demonstrates gender equality for the ability measured.
D) None of these

Accepted Answer

The fact that there are a large number of items biased in favor of one gender is concerning and indicates that the test may not be measuring the same ability in both groups. The equal number of biased items in favor of each gender does not cancel out the potential bias and does not guarantee fairness or gender equality. Therefore, none of the given choices are reasonable conclusions to draw from these findings.

Question 5

On a true/false inventory,a respondent selects true for an item that reads,"I summer in Tehran." The individual scoring the test would BEST interpret this response as indicative of the fact that this respondent
A) is extremely wealthy, with great holdings in oil.
B) is responding randomly to test items.
C) is of Middle Eastern descent.
D) requires more stimulation than Cape Cod has to offer.

Accepted Answer

The statement "I summer in Tehran" is clearly false, as Tehran is known for its hot and dry summers. Therefore, selecting "true" for this statement indicates that the respondent is not taking the test seriously and may be responding randomly. The other answer choices are not necessarily indicated by the respondent's answer.

Question 6

To ensure consistency in scoring,test developers have employed
A) anchor protocols.
B) resolvers.
C) revolvers.
D) Both a and b

Accepted Answer

The answer of To ensure consistency in scoring,test developers have...

Question 7

When a test is translated from one language in one culture to another language in another culture,______ can help ensure that the original test and the translated test are reasonably equivalent and tapping the same construct.
A) a translator
B) IRT
C) bi-lingual people who are experts on the two cultures
D) All of these

Accepted Answer

A translator can ensure the test is accurately translated, bi-lingual people who are experts on the two cultures can ensure cultural and linguistic equivalence, and IRT can help in comparing the psychometric properties of the original test and the translated test to ensure they are equivalent. Therefore, all of these options can help ensure that the original test and the translated test are reasonably equivalent and tapping the same construct.

Question 8

Jana takes a personality test administered by the "True Compatibility Dating Service." According to the personalized,computerized personality profile that results,Jana learns that her need for exhibitionism is much greater than her need for stability.Since the test analyzes data only with regard to Jane,and no other client of the dating service,it may be assumed that the test was scored using
A) a diagnostic model.
B) a cumulative model.
C) an ipsative model of scoring.
D) truly compatible models.

Accepted Answer

The fact that Jana's score was compared only to her own responses suggests that the test used an ipsative model of scoring. In this type of scoring, each individual's scores are analyzed within the context of their own responses rather than being compared to a normative group. This allows for a personalized and individualized profile, but limits the ability to make comparisons across individuals.

Question 9

A test item functions differently in one group of testtakers as compared to another group of testtakers known to have the same level of an underlying trait.This phenomenon is known as:
A) dysfunctional item syndrome.
B) DIF.
C) DIF item difference.
D) DIF item incongruity.

Accepted Answer

The phenomenon described in the question is known as Differential Item Functioning (DIF), where a test item functions differently for one group of test-takers compared to another group of test-takers with the same level of the underlying trait. The other answer choices are not commonly used or recognized terms in the field of testing and assessment.

Question 10

Instruments that contain items that function differentially
A) may have reduced validity.
B) may have inflated reliability.
C) are last to be banked in an item bank.
D) are informally referred to as "DIFFED"

Accepted Answer

Instruments that contain items that function differentially may have reduced validity because they may introduce bias into the measurement by favoring certain groups over others. This can result in inaccurate or incomplete measurements of the construct being studied.

Question 11

To increase the precision of a test,test developers may have to
A) increase the number of items.
B) increase the number of response options.
C) Both a and b
D) None of these

Accepted Answer

Increasing the number of items and response options can help increase the precision of a test. More items can provide a more comprehensive assessment of the construct being measured, while increased response options can help to differentiate between nuanced levels of the construct.

Question 12

At the end-point in the process of creating an item bank,the final version of a test may take the form of an instrument
A) suitable for computerized-adaptive testing.
B) that has all of the items "fixed."
C) Both a and b
D) None of these

Accepted Answer

The final version of a test created through the process of creating an item bank may take the form of an instrument suitable for computerized-adaptive testing, where the items are selected based on the examinee's responses, and all of the items are fixed. Therefore, both options A and B are correct.

Question 13

The reason latent-trait theory is so-named has to do with the presumption that
A) latent traits exist in males and females to the same degree.
B) whatever the test is measuring is multidimensional in nature.
C) the variable being measured is never directly measurable itself.
D) None of these

Accepted Answer

Latent-trait theory presumes that the variable being measured (the trait) is never directly measurable itself, hence the term latent. This means that the trait is inferred based on the individual's responses to various indicators or measures. The theory does not make any assumptions about the prevalence or expression of traits in males versus females (option A), nor does it suggest that the trait being measured is multidimensional (option B).

Question 14

A strategy for cheating on an examination entails one testtaker memorizing items and later recalling and reciting them for the benefit of a future testtaker.This cheating strategy may be countered by
A) a computer-tailored test administration to each testtaker.
B) a computer-randomized presentation of test items.
C) Both a and b
D) None of these

Accepted Answer

Both a and b would be effective in countering this strategy for cheating on an exam. A computer-tailored test administration to each testtaker would make it difficult for a future testtaker to share the exact same exam with another testtaker. Additionally, a computer-randomized presentation of test items would make it difficult for a future testtaker to know the exact questions that would appear on the exam. By using both of these strategies, the likelihood of a testtaker being able to cheat by memorizing and sharing specific exam items would be greatly diminished.

Question 15

On an item characteristic curve,the steeper the curve,
A) the more latent the trait is presumed to be.
B) the greater the item reliability.
C) the less the item discrimination.
D) the greater the item discrimination.

Accepted Answer

The steeper the curve on an item characteristic curve, the greater the item discrimination. This is because a steep curve indicates that individuals who score high on the underlying trait are more likely to answer the item correctly than individuals who score low on the trait, indicating that the item is effectively discriminating between high-ability and low-ability individuals. A flatter curve would indicate less discrimination, as individuals of different abilities would be more likely to answer the item similarly.

Question 16

The Rasch model offers a way to model the probability that
A) a person on the border of passing and failing a test will actually succeed at a particular criterion measure.
B) a person with X ability will be able to perform at a level of Y.
C) a test with a standard deviation of X will have a mean score of Y.
D) a test with X reliability will have Y validity.

Accepted Answer

The Rasch model is used in psychometrics for measuring abilities or attitudes and it predicts the probability that a person with a certain level of ability (X) will succeed in a task of a certain difficulty level (Y).

Question 17

When testing is conducted by means of a computer within a CAT context,it means that
A) a testtaker's response to one item may automatically trigger what item will be presented next.
B) testing may be terminated based on some pre-set number of consecutive item failures.
C) testing may be terminated based on some pre-set, maximum number of items being administered.
D) All of these

Accepted Answer

All of these options are possible in computerized adaptive testing (CAT). CAT algorithms use a test-taker's responses to previous items to determine which item to present next, based on the estimated ability level. Testing may be terminated early if the algorithm determines with a high degree of confidence what the test-taker's ability level is, or if a pre-set number of consecutive item failures has been reached. Additionally, tests may have a maximum number of items to administer, to avoid exhausting the test-taker or taking up too much time.

Question 18

Test scores measuring latent traits can,in theory at least,take on values ranging from
A) 0 to infinity.
B) negative infinity to positive infinity.
C) 0 to one million.
D) negative one million to positive one million.

Accepted Answer

Test scores measuring latent traits can potentially take on both positive and negative values, so the range for these scores would be from negative infinity to positive infinity.

Question 19

A test developer of multiple-choice ability tests reviews data from a recent test administration.She discovers that testtakers who scored very high on the test as a whole,all responded to item 13 with the same incorrect choice.Accordingly,the test developer
A) assumes that members of the high-scoring group are making some sort of unintended interpretation of item 13.
B) plans to interview members of the high-scoring group to understand the basis for their choice.
C) Both a and b
D) should remove item 13 from the test and place a note in its stead that reads: "Go to Item 14."

Accepted Answer

The test developer should consider both understanding the unintended interpretation (A) and interviewing the high-scoring group (B) to gain insights into why they chose the incorrect answer.

Question 20

The process of DIF analysis entails
A) scrutinizing item response curves for DIF items.
B) interviewing people from different cultures.
C) administering tests in different ways.
D) Both b and c

Accepted Answer

The process of DIF analysis involves examining item response curves to identify any differential functioning among different groups, such as those from different cultures. While interviews and varying test administration methods may also be part of cross-cultural assessment, they are not specific to DIF analysis.

Quiz 8: Test Development

Possible applications of IRT were discussed in your textbook.Which of the following is NOT one of those possible applications?

A math test developer is interested in deriving an index of the difficulty of the average item for his math test.As his consultant on test development,you advise him that this index could be obtained by:

With regard to item-discrimination indices,a d equal to -1 is

On a true/false inventory,a respondent selects true for an item that reads,"I summer in Tehran." The individual scoring the test would BEST interpret this response as indicative of the fact that this respondent

To ensure consistency in scoring,test developers have employed

When a test is translated from one language in one culture to another language in another culture,______ can help ensure that the original test and the translated test are reasonably equivalent and tapping the same construct.

A test item functions differently in one group of testtakers as compared to another group of testtakers known to have the same level of an underlying trait.This phenomenon is known as:

Instruments that contain items that function differentially

To increase the precision of a test,test developers may have to

At the end-point in the process of creating an item bank,the final version of a test may take the form of an instrument

The reason latent-trait theory is so-named has to do with the presumption that

A strategy for cheating on an examination entails one testtaker memorizing items and later recalling and reciting them for the benefit of a future testtaker.This cheating strategy may be countered by

On an item characteristic curve,the steeper the curve,

The Rasch model offers a way to model the probability that

When testing is conducted by means of a computer within a CAT context,it means that

Test scores measuring latent traits can,in theory at least,take on values ranging from

A test developer of multiple-choice ability tests reviews data from a recent test administration.She discovers that testtakers who scored very high on the test as a whole,all responded to item 13 with the same incorrect choice.Accordingly,the test developer

The process of DIF analysis entails