Brief Note
Confounding Problems in Multifactor AOV When Using Several Organismic Variables of Limited Reliability
Bibliographic Data
| ID | 8223286 |
|---|---|
| Authors | Paul A Games (Pennsylvania State University, corresponding author) |
| Year | 1975 |
| Volume | 12 |
| Issue | 2 |
| Pages | 225-232 |
| Publication date | 1975-03-01 |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | American Educational Research Journal (JOURNAL) |
| Journal identifiers | ISSN: 0002-8312 • E-ISSN: 1935-1011 |
| Publisher | American Educational Research Association (AERA) (PUBLISHER) |
| DOI | 10.3102/00028312012002225 |
| OpenAlex | W2000337025 |
| Language | EN |
| Citations received | 1 |
| References cited | 8 |
Organismic variables are variables that may be obtained by measurement of subjects, but that are not assignable (in the given investigation) to subjects. Mental age, I.Q., personality characteristics, past education, etc., are examples of organismic variables. One of the major characteristics of most such variables is that they possess limited reliability. This characteristic creates problems that do not exist when working with manipulatable variables (treatments) or variables defined by physical properties where the error of measurement is trivial. The basic reliability regression phenomenon is that the estimate of a true score, in deviation score units, is t= pA,ix where pl'i = the appropriate reliability coefficient and x is the observed deviation score. Thus extreme groups on an organismic variable, X1, will show differential regression when remeasured (when Pl'i is less than one). High scoring groups will regress down, while low
Confounding · Econometrics · Linear regression · Regression · Regression analysis · Reliability (semiconductor · Statistics · Variable (mathematics · Variables · Advanced Statistical Methods and Models · Mathematics · Psychology
| Citations per year | 0.03 |
|---|---|
| Citation span | 1989 - 1989 (1) |
| Highly cited | No |