Editing Psychological testing (section)

==Principles==
The development of a psychological test requires careful research. Some of the elements of test development involve the following:
:* ''Standardization'' - All procedures and steps must be conducted with consistency from one testing site/testing occasion to another. Examiner subjectivity is minimized (see objectivity next). Major standardized tests are normed on large try-out samples in order to understand what constitutes high, low, and intermediate scores.
:* ''Objectivity'' - Scoring such that subjective judgments and biases are minimized; scores are obtained in a similar manner for every test taker (see below).
:* ''Discrimination'' - Scores on a test should discriminate members of extreme groups; for example, each subscale of the original [[MMPI]] distinguished hospitalized patients suffering from mental illness and members of a well comparison group.<ref>{{Cite book |last1=Domino |first1=George |url=https://books.google.com/books?id=OiKau0aqtsYC&dq=discrimination+psychological+test&pg=PT34 |title=Psychological Testing: An Introduction |last2=Domino |first2=Marla L. |date=2006-04-24 |publisher=Cambridge University Press |isbn=978-1-139-45514-5 |pages=34+ |language=en}}</ref><ref>{{Cite book |last=Hogan |first=Thomas P. |url=https://books.google.com/books?id=4AEdvwEACAAJ |title=Psychological Testing: A Practical Introduction |publisher=John Wiley & Sons, Incorporated |year=2019 |isbn=978-1-119-50690-4 |pages=171+ |language=en}}</ref>
:* ''Test Norms'' - Part of the standardization of large-scale tests (see above). Norms help psychologists learn about individual differences. For example, a normed personality scale can help psychologists understand how some people are high in [[negative affectivity]] (NA) and others are low or intermediate in NA. With many psychoeducational tests, test norms allow educators and psychologists obtain an age- or grade-referenced percentile rank, for example, in reading achievement.
:* ''Reliability'' - Refers to test or scale consistency. It is important that individuals score about the same if they take a test and an alternate form of the test or if they take the same test twice, within a short time window. Reliability also refers to response consistency from test item to test item.
:* ''Validity'' - Refers to evidence that demonstrates that a test or scale measures what it is purported to measure.<ref Name = "Nunnally"/><ref>{{Cite book|title=Psychology and work today: An introduction to industrial and organizational psychology|last1=Schultz|first1=Duane P.|last2=Schultz|first2=Sydney Ellen|publisher=Prentice Hall|year=2010|isbn=978-0205683581|edition=10th|location=Upper Saddle River, N.J.|pages=99–102|oclc=318765451}}</ref>

=== Sample of behavior ===
The term ''sample of behavior'' refers to an individual's performance on tasks that have usually been prescribed beforehand. For example, a spelling test for middle school students cannot include all the words in the vocabularies of middle schoolers because there are thousands of words in their lexicon; a middle school spelling test must include only a sample of words in their vocabulary. The samples of behavior must be reasonably representative of the behavior in question. The samples of behavior that make up a paper-and-pencil test, the most common type of psychological test, are written into the test items. Total performance on the items produces a test score. A score on a well-constructed test is believed to reflect a [[psychological construct]] such as achievement in a school subject like vocabulary or mathematics knowledge, [[cognitive ability]], dimensions of [[personality]] such as introversion/extraversion, etc. Differences in test scores are thought to reflect individual differences in the construct the test is purported to measure.<ref Name = "Nunnally"/>