Term
|
Definition
| Displays the scores of a variable to show how frequently each value was obtained |
|
|
Term
|
Definition
| Consists of previous learning |
|
|
Term
|
Definition
| Consists of one's potential for learning and acquiring a skill |
|
|
Term
|
Definition
| Consists of problem solving, adapting to change, abstract thinking, and profiting from experience -- a mixture of achievement and aptitude |
|
|
Term
|
Definition
| Provides limited information concerning the general shape of the distribution (Does NOT provide information about differences in a specific interval) |
|
|
Term
|
Definition
| Indices of the central value or location of a frequency distribution with respect to the X axis |
|
|
Term
|
Definition
| Takes into account all scores; the most "democratic" of all methods |
|
|
Term
|
Definition
| Outliers have substantial influence on the outcome |
|
|
Term
|
Definition
| Not affected by skewed data or outliers |
|
|
Term
|
Definition
| All points are ignored except for those on the 50th percentile |
|
|
Term
|
Definition
| An "interesting" statistic; possibly relevant frequently |
|
|
Term
|
Definition
| Relies on classification; is a nominal level statistic |
|
|
Term
|
Definition
| The average squared deviations around the mean |
|
|
Term
|
Definition
| The square root of the average squared deviation around the mean |
|
|
Term
|
Definition
| More, less, or same as a normal distribution |
|
|
Term
|
Definition
| Difference between a score and the mean, divided by Standard Deviation |
|
|
Term
| What 2 pieces of information does the Z-Score give us about a raw score? |
|
Definition
Location (Above or below distribution mean?)
Distance (How far the raw score falls from the mean) |
|
|
Term
| How are Z-Scores different from T-Scores? |
|
Definition
Z-Score has a mean of 0 and a SD of 1
T-Score can range from 0-100 with 5 SD's within (10z+50) |
|
|
Term
|
Definition
| Points that dive the frequency distribution into equal fourths |
|
|
Term
|
Definition
| Scores bounded by the 25th and 75th percentile (middle 50%) |
|
|
Term
| What is norm and what is it used for? |
|
Definition
Performances by defined groups on particular tests
Used as a reference when evaluating and interpreting individual test scores |
|
|
Term
| What is norming and what is it used for? |
|
Definition
When you take a group of people and see how they perform on a test compared to others
Used when you want to see how a particular group did compared to another. |
|
|
Term
| What is standardization and what is it used for? |
|
Definition
| When you take an individual outside of a group and comapare them to how they did compared to a previous group. |
|
|
Term
| What are the five characteristics of a good theory? |
|
Definition
Explanatory power
Broad scope
Systematic
Fruitful
Parsimonious |
|
|
Term
|
Definition
| Graphical tool that helps illustrate linear and non-linear relationships |
|
|
Term
| Correlation Coefficient, and what should it NOT be confused with? |
|
Definition
The degree of variation in one variable estimated based on the knowledge about another's variable variation.
Should NOT be confused with Causation |
|
|
Term
|
Definition
| Perfect 1:1 correspondence between paired scores in the X and Y distributions |
|
|
Term
|
Definition
| Perfect 1:1 inverse correspondence between paired scores in the X and Y distributions |
|
|
Term
Principle of Least Squares
Relation to Regression line |
|
Definition
| The best-fitting line is the one with the smallest sum of squared residuals |
|
|
Term
|
Definition
Degree to which two variables vary together
Degree to which X&Y share a linear relationship
Degree to which points of two variables fall along a least squares regression line |
|
|
Term
| Principle of dilution in correlation |
|
Definition
| An imperfect 2:1 correspondence between X&Y |
|
|
Term
| Pearson Product Moment Correlation |
|
Definition
| Most commonly used index of correlation, appropriately used when both variables are continuous |
|
|
Term
|
Definition
| Difference between predicted value and actual value (The amount of Y "leftover") |
|
|
Term
| Standard Error of the Estimate |
|
Definition
| An index of the accuracy of prediction |
|
|
Term
| Relationship between SEofE and Residuals? |
|
Definition
| If highly accurate the differences between observed and predicted will be small because of the data points |
|
|
Term
| Coefficient of Alienation |
|
Definition
| A measurement of how things are NOT related |
|
|
Term
|
Definition
| Often occurs when a regression equation is calculated using one group of subjects and used to predict performance in another group of subjects |
|
|
Term
Restricted Range
What else is it called?
What does it lead to? |
|
Definition
Occurs when the range of variables becomes restricted
Truncation
Leads to reduced variance and may significantly alter strength of the correlation |
|
|
Term
| Purpose of using discriminant analysis |
|
Definition
Used to predict group membership
(those who passed vs. those who failed; how many of you passed and studied X amount of hours) |
|
|
Term
|
Definition
A data reduction technique that identifies underlying constructs for us
An exploratory technique for exploring the data and then making decisions |
|
|
Term
| Co-efficient of determination |
|
Definition
An index of:
- The amount of shared variance between variables X&Y
- The degree to which the variation scores on X "explains" (but not necessarily causes) the variation in scores on Y |
|
|
Term
| What is the regression formula? |
|
Definition
Y'=a+bX
Y' = Predicted Value of Y
a = The Intercept
b = the slope
X = the variable |
|
|
Term
|
Definition
The higher the correlation between x and y, the stronger the correlation
Therefore, the more we're able to predict with confidence, the less likely we are to make an error |
|
|
Term
|
Definition
| Consistency in measurement operations, as well as stability |
|
|
Term
|
Definition
|
|
Term
| What three components make up Classical Test Score Theory? |
|
Definition
| Observed score, true score, error |
|
|
Term
|
Definition
One data point within an entire test score distribution if parallel versions were administered many times
Also, the true score plus of minus the error that's there (assumes unsystematic error) |
|
|
Term
| How can error impact the observed score? |
|
Definition
Measurement error leads to a difference in the true and observed scores
As standard error of the estimate increases/decreases, so does the size of the observed score |
|
|
Term
| What does Kuder Richardson let us do? |
|
Definition
| Way to test if you can predict how you do within the same test. |
|
|
Term
| Mathematical formula for Kuder Richardson? |
|
Definition
|
|
Term
| Mathematical formula for Kuder Richardson? |
|
Definition
|
|
Term
| What does Cronbach's Alpha do? |
|
Definition
| Looks at how all your items get along; measures the degree of judges' agreement using *continuous* data |
|
|
Term
| How does Cronbach's Alpha measure internal consistency? |
|
Definition
Appropriate for dichotomous and non-dichotomous values
Based on the average inter-item covariances
Mean of all possible split-half correlations, corrected by the SB formula |
|
|
Term
| Name the three ways test reliability is usually estimated |
|
Definition
Time-retest reliability
Parallel forms reliability
Internal Consistency |
|
|
Term
|
Definition
- Evaluates error from the same test given on two administrations
- Only valuable when measuring stable characteristics
- Usually overestimates/inflates the true reliability |
|
|
Term
| Parallel/Alternate Forms reliability |
|
Definition
- Compares two equivalent forms measuring the same attribute
- Different items between forms with the same selection rules
- Equivalent content, difficulty, means, variances
- Different sections (split half reliability) |
|
|
Term
|
Definition
| - Examines homogeneity with which a test measures a construct |
|
|
Term
|
Definition
| When the first session influences the second |
|
|
Term
|
Definition
Test is given once and divided into separately scored halves
Halves are compared by correlating the two halves
Examines the stability in scores for two equivalent halves of the same test |
|
|
Term
| Spearman-Brown Prophecy Formula |
|
Definition
| Predicts the reliability of a whole test based on using only part of it; improves estimation of split-half reliability. |
|
|
Term
|
Definition
| Examines consistency among judges evaluating the same behavior |
|
|
Term
|
Definition
| Measure of agreement between two or more judges who each rate a set of objects using nominal data |
|
|
Term
| What does the standard error of measurement do? |
|
Definition
| Represents that there is some error in prediction. |
|
|
Term
| What types of irregularities might make reliability coefficients biased or invalid? |
|
Definition
- Inflated variance
- Constricted variance
- Restriction of range
- Outliers |
|
|
Term
| How can one address/improve reliability? |
|
Definition
- Increase the length of test
- Throw out items that run down the reliability
- Estimate the true correlation if the test did not have measurement error |
|
|
Term
| Purpose of factor and item analysis? |
|
Definition
- Factor analysis is used to ensure items measure the same thing
- Tests are more reliable if they are unidimensional
- One factor accounts for more of the variance than any other |
|
|
Term
| What example was given in class regarding reliability? |
|
Definition
|
|
Term
|
Definition
| Offers two alternatives for each item |
|
|
Term
Advantages of dichotomous format?
Disadvantages? |
|
Definition
Adv: Simplicity, easy to administer, easy and quick scoring, flexibility, absolute judgment
Disadv: Encourage memorization, truth often comes in shades of gray, 50% guessing probability |
|
|
Term
|
Definition
| Resembles dichotomous except each item has more than two alternatives |
|
|
Term
Advantages of polytomous format?
Disadv.? |
|
Definition
Adv: Easy to administer and score, familiar format, probability of correct answer is lower than dichotomous
Disadv: Harder to write, need good distractors for reliability |
|
|
Term
| What types of questions are "selected-response format"? |
|
Definition
|
|
Term
| Which three types of questions would be considered "constructed-response format?" |
|
Definition
Short answer
Fill in the blank
Essay |
|
|
Term
| What are the 2 major formats of summative scales? |
|
Definition
Cumulative scoring
Subscale scoring |
|
|
Term
| What scales most frequently use the Likert format? |
|
Definition
Multiple-point scales
Personality and attitude |
|
|
Term
| What are the primary differences between the Likert and Category formats? |
|
Definition
- Likert uses 5, 6, or 7 point scales; Category uses many-point scales
- Category is used to make more fine-grained discrimination |
|
|
Term
| In creating a category format, the use of what will reduce error variance? |
|
Definition
| Clearly defined anchor intervals (endpoints and midpoints) |
|
|
Term
| When does the category format begin to reduce reliability? |
|
Definition
| After using more than the 9 or 10 point maximum. |
|
|
Term
| What are four questions that should be asked when generating a pool of candidate test items? |
|
Definition
- What construct is the test designed to measure?
- What are the test's objectives?
- Who will use this test?
- Who will take this test? |
|
|
