Sattler Ch. 3 & 4 Flashcards
Normal Curve
Bell-shaped distribution representing many human characteristics.
Characterized by symmetry, mean = median = mode, and specific percentages of scores fall within standard deviations from the mean
Dispersion
Variability of scores within a set, indicating how much scores deviate from each other.
Range: Simplest measure, difference between highest and lowest scores, but sensitive to extreme values.
Variance: Measures variability around the mean, considering all scores, calculated as the average squared deviation from the mean.
Standard Deviation: Square root of variance, indicating average distance of scores from the mean, used extensively in testing and measurement.
Correlation
Indicates the degree and direction of relationship between two variables.
Positive Correlation: High scores on one variable associated with high scores on the other.
Negative Correlation: High scores on one variable associated with low scores on the other.
Correlation Coefficient: Ranges from -1.00 (perfect negative correlation) to +1.00 (perfect positive correlation), with 0 indicating no relationship.
types of Correlation Coefficients:
Pearson Product Moment Correlation Coefficient (r): Used under specific conditions including continuity of variables and normal distribution.
Spearman Rank-Difference Method: Suitable for ordinal data or when Pearson’s conditions aren’t met, using ranks instead of actual scores.
Correlational Strength Descriptions
20 to .29: Weak
.30 to .49: Moderately Weak
.50 to .69: Moderate
.70 to .79: Moderately Strong
.80 to .99: Strong
Context and Purpose in Correlation Interpretation:
mportance of sample size in determining significance and strength of correlation.
Variance accounted for by the correlation coefficient (e.g., .20^2 = 4%, .70^2 = 49%).
Effects of range restriction and measurement error on correlation values.
Outliers can significantly impact correlation, necessitating data accuracy checks.
Correlation vs. Causation:
Correlations indicate relationships, not causality.
Example: Climate and malaria incidence correlation does not imply climate causes malaria.
Mediator and Moderator Variables:
Mediators explain how or why effects occur.
Moderators explain when effects occur.
Importance in understanding relationships and drawing conclusions.
Correcting for Measurement Error
Correction for attenuation aims to estimate correlation if variables were perfectly reliable.
Corrected scores can inflate relationships and are not always interpretable
Coefficient of Determination (r^2):
Squared correlation coefficient indicates variance explained in one variable by another.
Example: r = .60 between intelligence scores and school grades means intelligence accounts for 36% of the variance in grades.
Indicates association, not causation, between variables.
Regression Analysis:
Predicts and explains variables using a linear equation, Ypred = bX + a, where Ypred is the predicted score, b is the slope, X is the known score, and a is the Y-intercept.
Standard Error of Estimate:
: Measures accuracy of predicted Y scores, indicating the average deviation of observed scores from predicted scores. Lower standard error suggests higher prediction accuracy.
Multiple Correlation
: Assesses the prediction accuracy of one variable based on two or more other variables, improving prediction accuracy but requiring larger sample sizes for stability.
Norm-Referenced Measurement
Compares an individual’s performance to a norm group or standardization sample, providing a context for understanding performance.
Types of Interpretation:
Criterion-referenced: Compares performance against a set standard or mastery level.
Standards-referenced: Evaluates performance in relation to defined standards or benchmarks.
Norm-referenced: Compares performance to that of a representative group, allowing for relative performance assessment.
Derived Scores in Norm-Referenced Assessment:
Transformed raw scores that indicate an individual’s standing relative to the norm group, facilitating comparisons across different measures.
Norm-Referenced Test Administration:
Requires standardized procedures to ensure fairness and reliability of comparisons across individuals.
Item Response Theory (IRT):
Uses item difficulty, discrimination, and guessing parameters to evaluate test items.
An item characteristic curve represents the probability of correctly answering an item across different ability levels.
Differential Item Functioning (DIF):
Assesses if test items function differently across diverse groups, highlighting potential biases.
Validity:
indicates whether a test measures what it claims to measure.
Essential for drawing appropriate conclusions from test results.
Types of Validity:
Content validity: Items represent the domain being assessed.
Face validity: The test appears valid at face value.
Construct validity: Measures the intended construct, including convergent and discriminant validity aspects.
Criterion-related validity: Test scores correlate with a specific outcome, including concurrent and predictive validity.
Test Utility and Predictive Power:
Practical value of a test in aiding decision-making.
Predictive power evaluates the accuracy of decisions based on test scores, considering true positive, false positive, false negative, and true negative classifications.
Meta-Analysis:
Integrates findings from multiple studies on similar topics using quantitative methods.
Useful for validity generalization, though conclusions may have limitations due to study or meta-analytic procedure flaws.
Factor Analysis:
Exploratory and confirmatory analyses identify underlying factors among variables.
Factors are derived to explain patterns of intercorrelations, divided into communality, specificity, and error variance.
Psychometric Concepts and Test Variability:
Understanding differences in test results due to various factors, including child characteristics, test conditions, and psychometric properties.
Standard Deviation increment disbursement of data.
Concluding Remarks on Reliability and Validity:
No instrument is perfectly reliable or valid.
Validity must be considered in relation to the test’s specific purposes.
Assessment scores can vary based on many factors, emphasizing the careful interpretation of results.
Percentile Rank
- Way to compare children, Peer.
- Not a percentage
Reliability and Validity
- Reliable: demonstrate consistence result
- Validity: does it measure what it supposed to measure
Measure of central tendency,How to calculate
Mean, Median and mode.
- Floor effect
- Ceiling effect
Item Gradient
Norm table Layout
- Age or grade equivalents scores
- Skills areas assessed
- Test content
Publication date
Sampling
Assessment tools can be outdated and doesn’t take account of linguistic and cultural background.
Have to be able to justify tools.
Classification systems:
Determine supports they may need. 13 categories
SLD, OHI, ED, ID, AUT, TBI,OR,SLI,MD,DHH,VI, deafness, deaf/blind, medically establish condition ( prior 3rd birthday, the 14th one for CA)
Brain trauma must be after birth.
ASEESSMENT PROCESS: Overview
- Establishing rapport: Distractions on testers may impact results.
- Administering test items
- Keeping testing material read
- Responding empathically to child
- Know when to stop.
Testing limits procedures
- Providing additional cues or aids
- Changing stimulus modality
- Eliminating time limits
For new testers Challenges are more of students with limits abilities
Eligibility cons
- Scape goats for behaviors
- Self-fulfillment prophecies
- Limited explanation for difficulties
- Stereotyping
- Social judgment
- Preoccupations with label
- Suggest static profile.
