Term
|
Definition
| characteristic which can take on different values in different circumstances |
|
|
Term
|
Definition
| a functional association of a real number with each elementary event in a defined sample space. A statistical concept that is more of a process |
|
|
Term
|
Definition
| process of association of numbers with observations |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| a characteristic of population |
|
|
Term
|
Definition
| a characteristic of a sample. measure of only a sample |
|
|
Term
|
Definition
| Each element of the population has equivalent probability of entering into the sample |
|
|
Term
|
Definition
| finite number or limit on possible set of answers |
|
|
Term
|
Definition
| sample chosen represents larger population |
|
|
Term
|
Definition
| Manipulated variable (predictor) explanation of the variance in the DV |
|
|
Term
|
Definition
| outcome variable (predicted) |
|
|
Term
|
Definition
| anything else besides variables examined and are independent and not measured |
|
|
Term
|
Definition
| a potentially influencing extraneous variable that varies systematically with the independent variable. A kind of systematic error. Cannot assume causality when confound is present |
|
|
Term
|
Definition
| Variations in data oproduced extraneous variables. Not a mistake |
|
|
Term
|
Definition
| introduced by influencing variable, perhaps associated with lack of validity on an individual subject/case basis. Canbe viewed as a type of bias |
|
|
Term
|
Definition
| Related to lack of precision and lack of reliability |
|
|
Term
|
Definition
| Categories, groupings. No specific order. I.e. ethnicity, gender |
|
|
Term
|
Definition
| measurement scale for values that can be assigned a greater than or less than value. A ranking system. Cannot determine degree to which one is better than the next. |
|
|
Term
|
Definition
| difference between values on the scale has the same meaning. scores can be added or subtracted and has a mean as a representative measure of central tendency. No zero point, or one is assigned. |
|
|
Term
|
Definition
| an interval scale with an absolute zero. Scale must be unipolar (only positive values, negative meaningless) or bipolar (both positive and negative values are meaningful) and can be continuous or discrete |
|
|
Term
|
Definition
| observed. often used to measure a latent variable |
|
|
Term
|
Definition
| intervening or underlying variable that cannot be measured |
|
|
Term
|
Definition
| repeatability or consistence in repetition. |
|
|
Term
|
Definition
| the scale measures the material it was designed to measure (its content area)covers most of the content area |
|
|
Term
|
Definition
| reflects the success of measures used to prediction and estimation. Two types: concurrent and predictive |
|
|
Term
|
Definition
| measures the degree to which a test correlates with another test designed to measure a similar outcome |
|
|
Term
|
Definition
| the degree to which the measure can predict some construct..gre's |
|
|
Term
|
Definition
| emprical and theoretical support for the interpretation of the constructs |
|
|
Term
|
Definition
| cleanness of outcome... can we say that X causes Y |
|
|
Term
|
Definition
| generalizability of a causal relationship |
|
|
Term
|
Definition
| the larger the N, the closer the X should be to p(x) |
|
|
Term
|
Definition
| experiment or process that can only result in one of two outcomes. with replacement |
|
|
Term
|
Definition
| mean of sampling distribution equals population paramenter being estimated |
|
|
Term
|
Definition
| with larger sample sizes, the likelihood that an observed statistic deviates much from a population parameter goes down. Essentially the law of large numbers. With larger sample sizes, statistics become better estimaters. |
|
|
Term
|
Definition
| if looking at two statistics of a sampling distribution (mean or median), the more efficient estimator will have the smaller sampling variance and smaller error (when estimating a population parameter) |
|
|
Term
|
Definition
| meu and sigma are sufficient estimaters for normal distributions. Statistics that we can estimate distribution from. |
|
|
Term
|
Definition
| as N increases, sigma squared and x bar approaches meu, the sampling distribution of the mean approaches normality as N gets large. Occurs regardless of shape of original distribution |
|
|
