Technical Foundations Flashcards

Which item response theory (IRT) model uses only item difficulty as its parameter?

Answer: Rasch (1PL) model

The Rasch model (one-parameter logistic model) estimates examinee ability based solely on item difficulty, assuming equal discrimination across items.

A normal distribution of test scores has a mean of 70 and a standard deviation of 10. Approximately what percentage of scores fall between 60 and 80?

Answer: 68%

In a normal distribution, approximately 68% of scores fall within one standard deviation above and below the mean.

Which type of language test requires examinees to produce unrestricted written or spoken responses scored by raters?

Answer: Constructed-response test

Constructed-response tests require examinees to produce their own responses, which are then evaluated by trained raters using scoring criteria.

What is the primary purpose of anchor items in equating across multiple test forms?

Answer: To link scores from different forms to a common scale

Anchor items appear on multiple test forms and are used to statistically link and equate the score scales of those forms.

Inter-rater reliability is BEST measured by which statistic when ratings are on an ordinal scale?

Answer: Cohen's kappa

Cohen's kappa measures agreement between two raters on categorical or ordinal data while correcting for agreement due to chance.

Which principle of language assessment states that a test should reflect real-world language use tasks?

Answer: Authenticity

Authenticity in language testing means that test tasks should closely resemble the target language use situations examinees encounter in real life.

What does the standard error of measurement (SEM) indicate about a test score?

Answer: The expected variability in an individual's scores across repeated administrations

The SEM quantifies the amount of error in an individual score, indicating how much that score might fluctuate if the person were tested again under the same conditions.