Treating Matching as a Variable in a Sociological Experiment
Bibliographic Data
| ID | 7598485 |
|---|---|
| Authors | J Milton Yinger, Kiyoshi Ikeda (0000-0003-4857-6976), Frank Laycock |
| Year | 1967 |
| Volume | 32 |
| Issue | 5 |
| Pages | 801 |
| Publication date | 1967-10-01 |
| Open Access | No |
| Type | ARTICLE |
| Venue | American Sociological Review (JOURNAL) |
| Journal identifiers | ISSN: 0003-1224 • E-ISSN: 1939-8271 |
| Publisher | SAGE Publications (PUBLISHER • US) |
| DOI | 10.2307/2092027 |
| OpenAlex | W2330343224 |
| Language | EN |
| Citations received | 3 |
| References cited | 1 |
the general equation can be simplified by taking the coefficient of the first kind of difference to be zero. This approach requires a definite theory as to the form of nonlinearity. For example, one has to decide whether to use absolute values of difference terms, second or higher degree polynomials, or other specific terms. Unfortunately our present verbal formulations are insufficiently specific to justify the selection of any one model in preference to others. Furthermore the models that I have used have all required that one make rather strong assumptions regarding the magnitudes of some of the coefficients. It would be advantageous if one could simply specify the signs of these coefficients together with some weak assumptions regarding their relative magnitudes. This is the strategy suggested by Hyman, but it does not appear that it will yield definitive results.39 In conclusion, none of the possible resolutions known to me seems to offer the hope of completely satisfactory solutions. In view of this, one should make use of these intriguing but rather complex theoretical formulations with the full realization that definitive results may not be obtainable. Certainly, when an investigator wishes to apply these theoretical orientations to his own research, he should first test the adequacy of a simple additive model involving only main effects of his individual status variables. If the additive model produces results that are almost as satisfactory as a model allowing for interaction, then the former simpler model is to be preferred
Matching (statistics · Sociology · Statistics · Variable (mathematics · Contemporary Sociological Theory and Practice · Mathematics · Qualitative Comparative Analysis Research · Social Power and Status Dynamics
| Citations per year | 0.06 |
|---|---|
| Citation span | 1975 - 2006 (32) |
| Highly cited | No |