The Trends of Research Methods in Japanese Sociology, 1952-2018
Tarohmaru, Hiroshi
19, September, 2019
0.1 Descriptives
Table h.1: Descriptive statistics
## journal year.c special sex
## Hyoron :834 1950's:113 general:1069 female: 211
## Sosioroji:549 60's :204 special: 314 male :1157
## 70's :184 NA's : 15
## 80's :194
## 90's :190
## 2000's:277
## 2010's:221
## status tenured method2 multipleAuthors
## students :394 Mode :logical theoretical :739 Mode :logical
## associate :218 FALSE:713 historical :183 FALSE:1340
## professor :197 TRUE :531 ethnographic:225 TRUE :43
## researcher:115 NA's :139 quantitative:236
## lecturer :105
## (Other) :215
## NA's :1390.2 Trends of Methods
Figure h.1: NUMBERS of papers by methods and year
Figure h.2: RATIOs of papers by methods and year
Figure h.??? RATIOs of some types of Papers within each year (moving average within three time points
Figure h.3: RATIO of methods within each year and each journal (moving average for three years)
Figure h.4: RATIO of methods within each year and article type (moving average for three years)
Figure h.5: RATIO of methods within each year and sex (moving average for three years)
Figure h.5: RATIO of methods within each year and status (moving average for three years)
0.3 Multinomial Logit Model
\[ \log \left(\frac{P_{ij}}{P_{i1}}\right) = \beta_{0j} + \beta_{1j} X_{i1} + \beta_{2j} X_{i2} + \cdots + \beta_{kj} X_{ik}, \] supposing \(P_{i1}\) is the probability that a paper, \(i\) employs theoretical methods, \(P_{ij} \ (j = 2, \ 3, \ 4)\) is the probability that a paper, \(i\) employs, historical, ethnographic, or quantitative methods, and \(X_{i1}, \ \ldots, \ X_{ik}\) are independent variables. Parameters are estimated with the maximum likelihood method.
| historical | ethnographic | quantitative | ||
|---|---|---|---|---|
| (Intercept) | -1.36*** | -1.01*** | -0.98*** | |
| (0.28) | (0.25) | (0.25) | ||
| Sosioroji | 0.69*** | 0.39* | 0.20 | |
| (0.21) | (0.20) | (0.20) | ||
| special | -1.37** | -0.56 | -0.52 | |
| (0.49) | (0.40) | (0.33) | ||
| multipleAuthors | 0.25 | 1.47** | 2.49*** | |
| (0.81) | (0.56) | (0.43) | ||
| male | -0.57* | -0.75*** | -0.69** | |
| (0.24) | (0.22) | (0.23) | ||
| tenured | -0.08 | -0.19 | 0.31 | |
| (0.22) | (0.22) | (0.19) | ||
| year | 0.38*** | 0.54*** | 0.30*** | |
| (0.06) | (0.06) | (0.06) | ||
| year squared | 0.12*** | 0.13*** | 0.11*** | |
| (0.03) | (0.03) | (0.03) | ||
| special X year | 0.04 | 0.15 | -0.39** | |
| (0.13) | (0.20) | (0.13) | ||
| special X year squared | 0.09 | -0.13 | -0.15* | |
| (0.07) | (0.10) | (0.07) | ||
| Deviance | 2665.00 | 2665.00 | 2665.00 | |
| Num. obs. | 1231 | 1231 | 1231 | |
| p < 0.001, p < 0.01, p < 0.05 | ||||
Fig h.6: Trends of methods predicted by the logit model, assuming paper is general, written by a single male tenured author on Sosioroji