Path Analysis
Bibliographic Data
| ID | 7078069 |
|---|---|
| Authors | William H Swatos (corresponding author) |
| Year | 2007 |
| Publication date | 2007-02-15 |
| Open Access | Yes |
| Type | OTHER |
| Venue | The Blackwell Encyclopedia of Sociology (JOURNAL) |
| Publisher | Wiley (PUBLISHER • GB) |
| DOI | 10.1002/9781405165518.wbeosp008 |
| OpenAlex | W4205557425 |
| ISBN | 9781405165518 |
| Language | EN |
| References cited | 2 |
Because of the difficulty in sociological research of conducting experiments that will yield valid and meaningful results, sociologists have looked for quasi‐experimental designs in which statistical measures can be interrelated in a logical sequence to suggest causality. Path analysis is the most widely used of such approaches. Path analysis combines both a theoretical (or logical) analysis with a statistical analysis, inasmuch as a logical argument must be introduced to suggest appropriate causal sequences. While some of these logical connections are obvious (gender obviously comes before marriage, for example), others are a matter of theoretical argument as part of a larger project (does image of God precede or follow religious affiliation or are there ongoing interaction effects between the two?). Through multiple and partial regression analysis, and statistical controls that attempt to match test and control populations as closely as possible, a researcher attempts to demonstrate both quantitatively and logically a pattern of causality that would be similar to what might result if an actual experimental design were able to have been put into place by providing estimates of the total direct and indirect effects of one variable on another. In most simple causal modeling, path coefficients are beta weights , which represent a measurement of changes in a dependent variable in terms of standard deviation units for each of the other variables (standardized regression coefficients) in series, creating adjusted slopes of the regression line which are comparable from one variable to the next, working backward from the dependent variable
Argument (complex analysis · Causality (physics · Econometrics · Logical conjunction · Path (computing · Path analysis (statistics · Regression · Regression analysis · Statistical hypothesis testing · Statistics · Variable (mathematics · Variables · Agricultural Innovations and Practices · Computer Science · Mathematics · Organic Food and Agriculture · Social and Intergroup Psychology
| Highly cited | No |
|---|