Psychological Assessment and Testing

Psychological Assessment

• Psychological assessment involves gathering and integrating psychology-related data to conduct a psychological evaluation, using tools like tests and procedures.
• Psychological tests are instruments designed to measure psychological variables.
• An item is a stimulus to which a person responds overtly.
• Content pertains to the test's focus, such as personality or IQ.
• Format refers to a test's structure, arrangement, layout, and administration method (computerized, paper-pencil, etc.).
• A score is a summary statement that reflects an evaluation of performance, task, interview, or behavior sample.
• Scoring involves assigning codes or meaning to performance, tasks, or behavior samples.
• Scales relate raw scores to a theoretical or empirical distribution.
• A cut score is a reference point used to divide data into classifications.

Gathering Information

• An interview is a method of gathering information through direct communication and reciprocal exchange, noting both verbal and nonverbal behavior like body language and eye contact.
• Interviews can be conducted via telephone or other devices.
• A portfolio is a collection of work products in various media.
• Case history data includes records, transcripts, and accounts that preserve archival information, often found in schools, hospitals, and criminal justice agencies.
• Case studies assemble case history data into an illustrative account.
• Behavioral observation monitors actions visually or electronically, recording quantitative and/or qualitative data, often in natural settings.
• Naturalistic observation takes place in a natural setting.

Potential Errors in Observation

• Reactivity refers to a reaction to being observed.
• Drift is the tendency to deviate from established behavioral observation definitions, developing one's own definitions.
• Contrast effect is when the same behavior is rated differently across repeated observations.
• Halo effect is the tendency to ascribe positive attributes independently of the observed behavior.

Computers in Testing

• Computers are used for test administration, scoring, and interpretation, via local or central processing.
• Score reports can be simple listings, extended with statistical analyses, interpretive with numerical and narrative statements, consultative for communication between professionals, or integrative using other collected data.
• CAPA (computer-assisted psychological assessment) refers to computer assistance in testing.
• CAT (computer-adaptive testing) tailors the test to the testtaker's ability.

Applications and Uses of Tests

• Psychological assessment is used for classification or diagnosis, description of strengths and weaknesses, and prediction of future behavior or treatment planning.
• The process begins with a referral question from sources like psychiatrists or teachers.
• Assessments typically contain one or more referral questions.
• It includes acquiring knowledge of the problem, selecting assessment tools, collecting data (using multiple sources for consistency), and data interpretation.
• Interpretation describes functioning, infers from data, and provides feedback to the client.
• Psychological assessment is applied in educational settings to identify special needs and evaluate learning, in geriatric settings to assess cognitive and adaptive functioning, and in clinical settings to screen for behavior problems.
• It also assists in counseling for personal improvement, in business and military settings for career decisions, and in government for professional credentialing.

Parties Involved in Testing

• Test developers create tests, adhering to ethical standards.
• Test users include clinicians, counselors, psychologists, HR personnel, and social workers.
• Testtakers are subjects of assessment.
• Society at large organizes individual differences.
• Test user qualification levels are tiered, ranging from Level A (basic tests) to Level C (advanced tests requiring doctoral-level training).

Test Administration

• Guidelines exist for preparation, administration, scoring, and record storage.
• Assessors must be familiar with test materials, procedures, and venue.
• Rapport is important to establish during testing.
• Safeguarding test protocols and obtaining permission are essential after testing.
• Assessments for people with disabilities should offer accommodations.
• Accommodation involves adapting the test or procedure to suit exceptional needs, while alternate assessment varies the measurement method.

Informative References

• Test catalogues contain test descriptions and technical information.
• Test manuals provide test development and information.
• Reference volumes offer details on each test, and journal articles provide reviews and psychometric studies.

Psychological Testing Defined

• Psychological testing measures psychology-related variables.
• Psychological testing encompasses all uses, applications, and concepts involved with psychological and educational tests.
• Psychometrics is the science of psychological measurement.
• Psychometric soundness refers to a test's consistency and accuracy.

Historical Context

• The Han Dynasty (206 B.C.E. to 220 C.E.) used test batteries for civil law and military affairs.
• Testing was also used in the Ming Dynasty to evaluate eligibility for public office (1368-1644 C.E.)
• Hippocrates categorized personality types based on fluids.
• Christian von Wolff envisioned psychology as a science with psychological measurement.
• Charles Darwin's The Origins of Species (1859) introduced survival of the fittest and individual differences.
• Hereditary Genius (1869) highlighted individual sensory and motor functioning differences.
• The 19th century saw J.E. Herbert use mathematical models for educational theories.
• Ernst Weber explored psychological thresholds, and G.T. Fechner devised a law relating sensation strength to stimulus intensity.
• Wilhelm Wundt is credited with founding psychology in 1879.

20th Century Testing Advancements

• Alfred Binet developed the Binet-Simon Scale (1905) in France.
• WWI led to standardized achievement tests, including the Army Alpha and Beta.
• David Wechsler created the Wechsler-Bellevue Intelligence Scale (1939).
• Personality tests blossomed between 1920-1940.
• The Rorschach inkblot test and Thematic Apperception Test were introduced.

Key Personality Testing in the Early-Mid 1900's

• Robert S. Woodworth developed the Personal Data Sheet for emotional stability.
• The Minnesota Multiphasic Personality Inventory (1943) used empirical methods.
• Factor analysis became a trend in personality test creation.
• Raymond B. Cattell introduced the 16 Personality Questionnaire.

Test Types and Considerations

• Individual tests are one-on-one and require an active administrator, while group tests can be administered to many, and may not require administrator presence.
• Ability tests measure skills in speed and accuracy, while achievement tests assess previous learning, aptitude tests predict future learning, and intelligence tests gauge problem-solving potential. These three measures are highly interrelated.
• Personality tests assess typical behavior, traits and dispositions. Structured personality tests provide self-report statements, with true/false answers, while projective tests use ambiguous stimuli with unclear responses.

Testing Uses

• Psychological testing is applied to decision-making, psychological research, and self-understanding and personal development.

Administration Factors

• The behavior of the examiner and the relationship with the examinee affects test scores.
• Examiners must build rapport.
• The race or ethnicity has little effect on test scores, in many cases.
• Different assessments require different levels of administrator training.
• Expectancy effects can alter data findings.
• Reinforcement influences behavior.
• Computer-assisted testing may lead to less disclosure of socially undesirable information.
• Studies have shown that interviewer administered evaluations are more likely to find a person in good health than a self-completed test.
• Subject variables, such as test anxiety, can be a source of errors.

Testing vs. Assessment

• Testing obtains numerical gauges of abilities, while assessment answers referral questions and solves problems.
• Testing is individual or group-based focusing on results rather than processing. Assessment is typically individualized.
• Evaluators in testing are interchangeable, while assessors are key in assessment.
• Testing requires technical skills, while assessment requires educated skill and thoughtful data integration.
• Testing yields test scores, while assessment entails logical problem-solving, and has the intention of shedding light on a question raised in the referral.

Measurement and Variables

• Measurement assigns numbers or symbols to characteristics using rules.
• Discrete variables have finite values while continuous variables have infinite values.
• Error is the influence on a test score beyond what's measured.
• A scale is a set of numbers modelling properties.

Scales of Measurement

• Nominal scales categorize, and allow counting frequencies inside the category.
• Ordinal scales rank order, like birth order, but do not measure distance between positions, and can be statistically manipulated like nominal scales.
• Interval scales have magnitude and equal intervals, like IQ scores.
• Ratio scales have magnitude, equal intervals, and an absolute zero.

Norms, Statistics, and Distributions

• Parameters are measures derived from populations; statistics are measures from samples.
• Descriptive statistics describe quantitative information, while inferential statistics make inferences from samples to populations.
• A distribution is a set of test scores arrayed for recording or study.
• Raw scores are unmodified performance accounts.

Frequency, Distributions and Measures

• Frequency distributions list scores with their occurrence, in both tabular and graph form
• Measures of central tendency indicate the average score in a distribution.
• Arithmetic mean is affected by the actual numerical value of every score
• Median is the middle score.
• Mode is the most frequent score.
• Multimodal data will have 2+ scores occurring with equal frequency.

Assessing Variability

• Measures of variability describe the amount of variation.
• Range is the difference between the highest and lowest scores.
• Interquartile range divides scores into four parts.
• Average deviation is the average distance between values and the mean.
• Standard deviation is the square root of the averaged squared deviations.
• Skewness is the extent to which symmetry is absent.
• Kurtosis is the steepness of the distribution.

The Normal Curve

• Abraham DeMoivre, and Karl Friedrich Gauss contributed to the idea of the normal curve.
• The normal curve is bell-shaped, smooth, and mathematically defined.
• Standard scores are raw scores converted to a scale with an arbitrarily set mean and standard deviation
• A Z score has a mean of 0 and a standard deviation of 1.
• A T score has a mean of 50 and a standard deviation of 10.
• Stanine converts scores into a scale from 1 to 9.

Norms and Test Interpretation

• Norm-referenced testing derives meaning from test scores by comparing them to a group of testtakers.
• The sample is called a normative group.
• Standardization involves administering a test to a representative sample and establishing norms.
• Norms can be percentiles, age/grade equivalents, or developmental.

Normative Groups

• Norms are based on trait, ability or skill which are affected by age/grade, or stage of life.
• Norms can be national, subgroup, or local.
• Criterion-referenced testing evaluates scores against a set standard.

Correlation, Variables and Statistical Relationships

• Correlation is a graph of coordinate pairs for X and Y variables, and the eyeball gauge of its curvature
• Scatterplots is a quick indication of the direction and magnitude of how two variables are related.
• An outlier is unrepresentative and distant from coordinate points.
• Correlation is a mathematical index that describes the direction and magnitude of a relationship.

Types of correlations

• Positive correlation means high on Y are associated with high scores on X; low on Y corresponds to low on X.
• Negative correlations mean high on Y is associated with inverse scores on X.
• No correlation means that two variables are not related.

Correlation Calculations

• Pearson r measures continuous linear relationships.
• Spearman's rho measures sets of ranks.
• Biserial correlation reports between a continuous and artificial variable.
• Point-biserial measures the correlation between true dichotomous and continuous variables.
• Phi coefficient measures the relationship between two true dichotomous, or true and artificial dichotomous variables.
• Tetrachoric measures relationships between 2 artificial variables.
• Regression is used for predicting scores.

Errors in data prediction and analysis

• Correlation alone cannot demonstrate correlation because there might be a third variable
• Residual is the difference between observed and predicted scores.
• Restricted range problems occur if the variability is restricted, leading to difficult correlations

Reliability and Validity

• Reliability means the consistency of data taken, and dependabilty.
• Test construction errors, test administration errors, or test scoring and interpretation might affect reliability of testing.
• Estimating reliability is performed through a test-retest, via parallel or alternate administering, or split half via correlating two pairs of equivalent halves

Calculations of Validity

• Spearman-Brown formula estimates internal consistency from correlating two test halves.
• Kuder-Richardson formula 20 estimates inter-item consistency within homogenous and dichotomous test.
• Coefficient Alpha estimates internal consistency of tests that are non-dichotomous, and preferred statistical method for measuring consistency

Inter-rater Reliability

• inter-score reliability measures the degree of consistency from multiple test-raters, assessed via Kappa statistical analysis
• Values of Kappa vary in values between +1 and -1.
• Standard error of measurement (SEM) estimates the degree of which a score may deviate. SEM and reliability are inversely proportional.