#
dta <- read.csv("data0419.csv", header = T)
options(digits = 3)
pacman::p_load(tidyverse, ggplot2)
dta <- dta %>% mutate( Gender = relevel(Gender, ref = "女"),
Sector = relevel(Sector, ref = "私立"),
Field1 = relevel(Field1, ref = "遊憩與運動學群"),
EduLv = factor(EduLv, levels=c("博士","碩士","普通大學","科技大學",
"技術學院","五專","三專",
"二專","高中","高職","國中")),
EduLv = relevel(EduLv, ref = "技術學院"),
Region = factor(Region, levels =c("宜花東離島","北北基","桃竹苗",
"中彰投","雲嘉南","高屏澎")),
Age = as.numeric(Age),
J_year = as.numeric(J_year),
JobZone = as.numeric(JobZone),
EduZone = as.numeric(EduZone),
JobZone_D = as.numeric(EduZone-JobZone),
Salary = as.numeric(Salary),
SubEduOver = relevel(SubEduOver, ref="符合工作要求"),
Core = recode_factor(as.factor(JobCor), "1" = "無關聯",
"2" = "部分關聯",
"3" = "核心關聯"),
SubEduOver = factor(SubEduOver,levels =c("符合工作要求","高於工作要求","低於工作要求")))%>%
filter(Age >= 20)
# data construction
glimpse(dta)## Observations: 1,568
## Variables: 25
## $ SID <fctr> A10, A100, A103, A104, A105, A106, A107, A108, A1...
## $ Gender <fctr> 女, 女, 男, 女, 女, 女, 女, 女, 女, 男, 男, 男, 女, 男, 男, 女, 女...
## $ Sector <fctr> 國立（公立）, 私立, 國立（公立）, 國立（公立）, 國立（公立）, 私立, 私立, 國立（公立...
## $ EduLv <fctr> 碩士, 普通大學, 高職, 普通大學, 普通大學, 普通大學, 普通大學, 普通大學, 碩士, 碩...
## $ SubEduOver <fctr> 高於工作要求, 符合工作要求, 符合工作要求, 符合工作要求, 符合工作要求, 符合工作要求, 符...
## $ Require <fctr> 高中/高職, 普通大學, 高中/高職, 普通大學, 普通大學, 普通大學, 普通大學, 普通大學,...
## $ Field1 <fctr> 資訊學群, 外語學群, 工程學群, 文史哲學群, 文史哲學群, 大眾傳播學群, 大眾傳播學群, 藝...
## $ City <fctr> 高雄市, 苗栗縣, 高雄市, 南投縣, 嘉義市, 臺北市, 臺北市, 南投縣, 高雄市, 臺中市,...
## $ Category <fctr> 受雇於公營機關, 受雇於公營機關, 受雇者於私營企業, 受雇於公營機關, 受雇者於私營企業, 受雇...
## $ Staff <fctr> 50-99人, 50-99人, 2-9人, 100-199人, 10-29人, 30-49人, 3...
## $ Hours <int> 40, 70, 57, 51, 64, 50, 50, 47, 50, 60, 45, 40, 56...
## $ J_year <dbl> 8, 4, 21, 1, 6, 0, 1, 1, 17, 7, 3, 23, 1, 2, 1, 1,...
## $ J_total <dbl> 8, 4, 30, 1, 6, 0, 2, 2, 28, 7, 30, 26, 1, 10, 1, ...
## $ income <fctr> 2-3萬以下, 3-4萬以下, 3-4萬以下, 4-5萬以下, 2萬以下, 3-4萬以下, 3-4...
## $ SubMismatch <int> 2, 3, 5, 4, 5, 4, 3, 3, 4, 4, 4, 4, 5, 5, 3, 2, 4,...
## $ JobSat <int> 4, 3, 5, 6, 7, 5, 3, 6, 3, 5, 4, 7, 3, 4, 4, 4, 5,...
## $ EduZone <dbl> 5, 4, 2, 4, 4, 4, 4, 4, 5, 5, 3, 5, 4, 4, 4, 4, 5,...
## $ Region <fctr> 高屏澎, 桃竹苗, 高屏澎, 中彰投, 雲嘉南, 北北基, 北北基, 中彰投, 高屏澎, 中彰投,...
## $ Salary <dbl> 25000, 35000, 35000, 45000, 20000, 35000, 35000, 4...
## $ Age <dbl> 34, 30, 62, 25, 21, 24, 25, 26, 57, 35, 54, 54, 23...
## $ JobZone <dbl> 3, 4, 3, 4, 4, 4, 4, 4, 4, 4, 4, 5, 4, 4, 3, 2, 4,...
## $ JobCor <int> 1, 2, 1, 2, 1, 3, 1, 1, 1, 3, 1, 1, 3, 1, 3, 1, 2,...
## $ Core <fctr> 無關聯, 部分關聯, 無關聯, 部分關聯, 無關聯, 核心關聯, 無關聯, 無關聯, 無關聯, 核...
## $ ObjOver <fctr> over, adequate, under, adequate, adequate, adequa...
## $ JobZone_D <dbl> 2, 0, -1, 0, 0, 0, 0, 0, 1, 1, -1, 0, 0, 0, 1, 2, ...# NA
apply(apply(dta, 1, is.na), 1, sum)## SID Gender Sector EduLv SubEduOver Require
## 0 0 0 0 0 0
## Field1 City Category Staff Hours J_year
## 0 0 0 0 0 0
## J_total income SubMismatch JobSat EduZone Region
## 0 0 0 0 0 0
## Salary Age JobZone JobCor Core ObjOver
## 0 0 0 0 0 0
## JobZone_D
## 0# check and pick out
lapply(dta[,c("Sector", "Field1", "City", "Region","EduLv", "SubEduOver", "ObjOver")], levels)## $Sector
## [1] "私立" "國外學校" "國立（公立）"
##
## $Field1
## [1] "遊憩與運動學群" "大眾傳播學群" "工程學群" "文史哲學群"
## [5] "外語學群" "生命科學學群" "生物資源學群" "地球與環境學群"
## [9] "法政學群" "社會與心理學群" "建築與設計學群" "財經學群"
## [13] "教育學群" "資訊學群" "管理學群" "數理化學群"
## [17] "醫藥衛生學群" "藝術學群"
##
## $City
## [1] "宜蘭縣" "花蓮縣" "金門縣" "南投縣" "屏東縣" "苗栗縣" "桃園市"
## [8] "高雄市" "基隆市" "雲林縣" "新北市" "新竹市" "新竹縣" "嘉義市"
## [15] "嘉義縣" "彰化縣" "臺中市" "臺北市" "臺東縣" "臺南市" "澎湖縣"
##
## $Region
## [1] "宜花東離島" "北北基" "桃竹苗" "中彰投" "雲嘉南"
## [6] "高屏澎"
##
## $EduLv
## [1] "技術學院" "博士" "碩士" "普通大學" "科技大學" "五專"
## [7] "三專" "二專" "高中" "高職" "國中"
##
## $SubEduOver
## [1] "符合工作要求" "高於工作要求" "低於工作要求"
##
## $ObjOver
## [1] "adequate" "over" "under"names(dta)## [1] "SID" "Gender" "Sector" "EduLv" "SubEduOver"
## [6] "Require" "Field1" "City" "Category" "Staff"
## [11] "Hours" "J_year" "J_total" "income" "SubMismatch"
## [16] "JobSat" "EduZone" "Region" "Salary" "Age"
## [21] "JobZone" "JobCor" "Core" "ObjOver" "JobZone_D"p <- dplyr::select(dta, -City, -income, -JobSat)對照組設定：
女、私立、遊憩運動學群、技術學院、宜花東離島、過量教育(符合工作要求)、客評關聯（無關聯）
分別看年齡與工作年資的效果
#
lm1 <- lm(log(Salary) ~ Gender + Age + Sector + Region + Hours, data = p)
summary(lm1)##
## Call:
## lm(formula = log(Salary) ~ Gender + Age + Sector + Region + Hours,
## data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.0732 -0.2324 -0.0141 0.2017 1.7082
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 9.581178 0.078394 122.22 < 2e-16 ***
## Gender男 0.180206 0.018919 9.52 < 2e-16 ***
## Age 0.022415 0.001277 17.55 < 2e-16 ***
## Sector國外學校 0.321423 0.071735 4.48 8.0e-06 ***
## Sector國立（公立） 0.168267 0.018853 8.93 < 2e-16 ***
## Region北北基 0.047873 0.051154 0.94 0.3495
## Region桃竹苗 0.150990 0.054430 2.77 0.0056 **
## Region中彰投 -0.057720 0.053995 -1.07 0.2852
## Region雲嘉南 -0.061314 0.053261 -1.15 0.2498
## Region高屏澎 -0.095102 0.053134 -1.79 0.0737 .
## Hours 0.004167 0.000953 4.37 1.3e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.362 on 1557 degrees of freedom
## Multiple R-squared: 0.295, Adjusted R-squared: 0.291
## F-statistic: 65.2 on 10 and 1557 DF, p-value: <2e-16#
lm11 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours , data = p)
summary(lm11)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9217 -0.2546 -0.0138 0.2134 1.8340
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.137618 0.070260 144.29 < 2e-16 ***
## Gender男 0.194683 0.019515 9.98 < 2e-16 ***
## J_year 0.023136 0.001684 13.74 < 2e-16 ***
## Sector國外學校 0.368650 0.074000 4.98 7.0e-07 ***
## Sector國立（公立） 0.176812 0.019500 9.07 < 2e-16 ***
## Region北北基 0.060635 0.052915 1.15 0.2520
## Region桃竹苗 0.171719 0.056311 3.05 0.0023 **
## Region中彰投 -0.045241 0.055835 -0.81 0.4179
## Region雲嘉南 -0.056319 0.055048 -1.02 0.3064
## Region高屏澎 -0.073569 0.054904 -1.34 0.1805
## Hours 0.004150 0.000985 4.21 2.7e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.374 on 1557 degrees of freedom
## Multiple R-squared: 0.247, Adjusted R-squared: 0.242
## F-statistic: 51.1 on 10 and 1557 DF, p-value: <2e-16#
lm21 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours+ EduLv , data = p)
summary(lm21)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + EduLv, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9267 -0.2229 -0.0171 0.1873 1.8520
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.10554 0.08561 118.04 < 2e-16 ***
## Gender男 0.18835 0.01853 10.17 < 2e-16 ***
## J_year 0.02384 0.00169 14.14 < 2e-16 ***
## Sector國外學校 0.14817 0.07201 2.06 0.03979 *
## Sector國立（公立） 0.07792 0.01999 3.90 0.00010 ***
## Region北北基 0.04135 0.04988 0.83 0.40727
## Region桃竹苗 0.11183 0.05328 2.10 0.03600 *
## Region中彰投 -0.05160 0.05262 -0.98 0.32700
## Region雲嘉南 -0.08297 0.05200 -1.60 0.11079
## Region高屏澎 -0.07421 0.05180 -1.43 0.15217
## Hours 0.00361 0.00093 3.88 0.00011 ***
## EduLv博士 0.54772 0.12763 4.29 1.9e-05 ***
## EduLv碩士 0.32912 0.05962 5.52 4.0e-08 ***
## EduLv普通大學 0.09173 0.05786 1.59 0.11307
## EduLv科技大學 -0.00558 0.06011 -0.09 0.92607
## EduLv五專 0.13609 0.08494 1.60 0.10931
## EduLv三專 -0.13029 0.35805 -0.36 0.71599
## EduLv二專 0.03531 0.08222 0.43 0.66769
## EduLv高中 0.18480 0.25631 0.72 0.47103
## EduLv高職 -0.21464 0.08079 -2.66 0.00797 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.353 on 1548 degrees of freedom
## Multiple R-squared: 0.336, Adjusted R-squared: 0.328
## F-statistic: 41.2 on 19 and 1548 DF, p-value: <2e-16#
lm22 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours+ Field1 , data = p)
summary(lm22)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + Field1, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9261 -0.2478 -0.0178 0.2127 2.0052
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.057216 0.093736 107.29 < 2e-16 ***
## Gender男 0.149272 0.021941 6.80 1.5e-11 ***
## J_year 0.022387 0.001664 13.45 < 2e-16 ***
## Sector國外學校 0.379724 0.072981 5.20 2.2e-07 ***
## Sector國立（公立） 0.185438 0.020664 8.97 < 2e-16 ***
## Region北北基 0.081320 0.052310 1.55 0.12026
## Region桃竹苗 0.166556 0.055840 2.98 0.00290 **
## Region中彰投 -0.021551 0.055114 -0.39 0.69582
## Region雲嘉南 -0.049608 0.054263 -0.91 0.36075
## Region高屏澎 -0.072164 0.054159 -1.33 0.18291
## Hours 0.003476 0.000974 3.57 0.00037 ***
## Field1大眾傳播學群 0.027006 0.083860 0.32 0.74746
## Field1工程學群 0.239929 0.068770 3.49 0.00050 ***
## Field1文史哲學群 0.015746 0.075661 0.21 0.83517
## Field1外語學群 0.099669 0.075549 1.32 0.18727
## Field1生命科學學群 0.106513 0.089361 1.19 0.23347
## Field1生物資源學群 0.056530 0.095606 0.59 0.55442
## Field1地球與環境學群 0.062680 0.098800 0.63 0.52590
## Field1法政學群 0.097547 0.079010 1.23 0.21716
## Field1社會與心理學群 0.092296 0.071828 1.28 0.19900
## Field1建築與設計學群 -0.029695 0.079388 -0.37 0.70842
## Field1財經學群 0.095413 0.072192 1.32 0.18648
## Field1教育學群 0.123934 0.074014 1.67 0.09424 .
## Field1資訊學群 0.146845 0.072048 2.04 0.04171 *
## Field1管理學群 0.033969 0.070390 0.48 0.62946
## Field1數理化學群 0.177664 0.079138 2.25 0.02491 *
## Field1醫藥衛生學群 0.278198 0.072985 3.81 0.00014 ***
## Field1藝術學群 -0.022692 0.086327 -0.26 0.79270
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.367 on 1540 degrees of freedom
## Multiple R-squared: 0.286, Adjusted R-squared: 0.273
## F-statistic: 22.8 on 27 and 1540 DF, p-value: <2e-16客觀只有線縮教育的部分顯著>< 不~
自評的效果倒是很好
# 客觀
lm31 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours +ObjOver , data = p)
summary(lm31)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + ObjOver, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9224 -0.2506 -0.0126 0.2168 1.8321
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.121816 0.071124 142.31 < 2e-16 ***
## Gender男 0.192688 0.019500 9.88 < 2e-16 ***
## J_year 0.022876 0.001691 13.53 < 2e-16 ***
## Sector國外學校 0.373561 0.073977 5.05 4.9e-07 ***
## Sector國立（公立） 0.182506 0.019671 9.28 < 2e-16 ***
## Region北北基 0.058555 0.052840 1.11 0.268
## Region桃竹苗 0.167419 0.056314 2.97 0.003 **
## Region中彰投 -0.050737 0.055808 -0.91 0.363
## Region雲嘉南 -0.063710 0.055037 -1.16 0.247
## Region高屏澎 -0.079604 0.054868 -1.45 0.147
## Hours 0.004058 0.000985 4.12 4.0e-05 ***
## ObjOverover 0.026216 0.020879 1.26 0.209
## ObjOverunder 0.103931 0.040371 2.57 0.010 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.374 on 1555 degrees of freedom
## Multiple R-squared: 0.25, Adjusted R-squared: 0.245
## F-statistic: 43.3 on 12 and 1555 DF, p-value: <2e-16# 自評
lm32 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + SubEduOver , data = p)
summary(lm32)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + SubEduOver, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9488 -0.2389 -0.0265 0.2032 1.7671
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.186977 0.069174 147.27 < 2e-16 ***
## Gender男 0.206814 0.019246 10.75 < 2e-16 ***
## J_year 0.022222 0.001659 13.40 < 2e-16 ***
## Sector國外學校 0.400140 0.072814 5.50 4.5e-08 ***
## Sector國立（公立） 0.171572 0.019307 8.89 < 2e-16 ***
## Region北北基 0.043488 0.051931 0.84 0.4025
## Region桃竹苗 0.156990 0.055253 2.84 0.0046 **
## Region中彰投 -0.057601 0.054772 -1.05 0.2931
## Region雲嘉南 -0.068673 0.054003 -1.27 0.2037
## Region高屏澎 -0.080964 0.053854 -1.50 0.1329
## Hours 0.004493 0.000967 4.64 3.7e-06 ***
## SubEduOver高於工作要求 -0.154841 0.023760 -6.52 9.7e-11 ***
## SubEduOver低於工作要求 -0.174413 0.030482 -5.72 1.3e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.367 on 1555 degrees of freedom
## Multiple R-squared: 0.277, Adjusted R-squared: 0.271
## F-statistic: 49.7 on 12 and 1555 DF, p-value: <2e-16無論是客觀的關聯還是自評的關聯程度效果都很不錯
## 關聯
# 客觀
lm41 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + Core , data = p)
summary(lm41)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + Core, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9703 -0.2486 -0.0198 0.2075 1.8821
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.085836 0.069678 144.75 < 2e-16 ***
## Gender男 0.201106 0.019264 10.44 < 2e-16 ***
## J_year 0.023695 0.001664 14.24 < 2e-16 ***
## Sector國外學校 0.382650 0.073025 5.24 1.8e-07 ***
## Sector國立（公立） 0.174247 0.019282 9.04 < 2e-16 ***
## Region北北基 0.063799 0.052168 1.22 0.2215
## Region桃竹苗 0.177935 0.055523 3.20 0.0014 **
## Region中彰投 -0.034378 0.055067 -0.62 0.5325
## Region雲嘉南 -0.051353 0.054276 -0.95 0.3442
## Region高屏澎 -0.069485 0.054128 -1.28 0.1994
## Hours 0.003985 0.000973 4.10 4.4e-05 ***
## Core部分關聯 0.131083 0.028462 4.61 4.4e-06 ***
## Core核心關聯 0.133151 0.022243 5.99 2.7e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.369 on 1555 degrees of freedom
## Multiple R-squared: 0.269, Adjusted R-squared: 0.264
## F-statistic: 47.8 on 12 and 1555 DF, p-value: <2e-16# 自評
lm42 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + SubMismatch , data = p)
summary(lm42)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + SubMismatch, data = p)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.0169 -0.2414 -0.0095 0.1995 1.6998
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 9.931681 0.071733 138.45 < 2e-16 ***
## Gender男 0.206495 0.019025 10.85 < 2e-16 ***
## J_year 0.022649 0.001639 13.82 < 2e-16 ***
## Sector國外學校 0.365808 0.071985 5.08 4.2e-07 ***
## Sector國立（公立） 0.163094 0.019024 8.57 < 2e-16 ***
## Region北北基 0.057717 0.051475 1.12 0.2623
## Region桃竹苗 0.170214 0.054777 3.11 0.0019 **
## Region中彰投 -0.042604 0.054315 -0.78 0.4329
## Region雲嘉南 -0.058652 0.053549 -1.10 0.2736
## Region高屏澎 -0.068784 0.053411 -1.29 0.1980
## Hours 0.003976 0.000959 4.15 3.5e-05 ***
## SubMismatch 0.070426 0.007448 9.46 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.364 on 1556 degrees of freedom
## Multiple R-squared: 0.288, Adjusted R-squared: 0.283
## F-statistic: 57.2 on 11 and 1556 DF, p-value: <2e-16客觀過量教育一樣只有限縮的部份有顯著，客觀關聯效果高顯著
自評過量、關聯效果一樣不錯><
#
p1 <- p[ p$EduLv %in%
c("博士","碩士", "普通大學", "科技大學", "技術學院", "五專","三專","二專"), ]
#
lmn1 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + ObjOver , data = p1)
summary(lmn1)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + ObjOver, data = p1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9261 -0.2488 -0.0159 0.2117 1.8417
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.144147 0.071848 141.19 < 2e-16 ***
## Gender男 0.196379 0.019591 10.02 < 2e-16 ***
## J_year 0.024592 0.001756 14.01 < 2e-16 ***
## Sector國外學校 0.364167 0.073345 4.97 7.6e-07 ***
## Sector國立（公立） 0.185631 0.019808 9.37 < 2e-16 ***
## Region北北基 0.050767 0.053284 0.95 0.34086
## Region桃竹苗 0.171319 0.056801 3.02 0.00260 **
## Region中彰投 -0.053554 0.056257 -0.95 0.34127
## Region雲嘉南 -0.062325 0.055650 -1.12 0.26291
## Region高屏澎 -0.077265 0.055407 -1.39 0.16337
## Hours 0.003727 0.000991 3.76 0.00018 ***
## ObjOverover 0.011764 0.020944 0.56 0.57442
## ObjOverunder 0.136029 0.043430 3.13 0.00177 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.37 on 1515 degrees of freedom
## Multiple R-squared: 0.26, Adjusted R-squared: 0.254
## F-statistic: 44.4 on 12 and 1515 DF, p-value: <2e-16#
lmn2 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + Core , data = p1)
summary(lmn2)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + Core, data = p1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9643 -0.2467 -0.0187 0.2050 1.8851
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.102793 0.070633 143.03 < 2e-16 ***
## Gender男 0.203289 0.019391 10.48 < 2e-16 ***
## J_year 0.025279 0.001736 14.57 < 2e-16 ***
## Sector國外學校 0.370316 0.072577 5.10 3.8e-07 ***
## Sector國立（公立） 0.173668 0.019426 8.94 < 2e-16 ***
## Region北北基 0.058845 0.052706 1.12 0.2644
## Region桃竹苗 0.182187 0.056096 3.25 0.0012 **
## Region中彰投 -0.034615 0.055600 -0.62 0.5337
## Region雲嘉南 -0.047759 0.054967 -0.87 0.3851
## Region高屏澎 -0.064764 0.054760 -1.18 0.2371
## Hours 0.003657 0.000981 3.73 0.0002 ***
## Core部分關聯 0.127871 0.028568 4.48 8.2e-06 ***
## Core核心關聯 0.126141 0.022216 5.68 1.6e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.366 on 1515 degrees of freedom
## Multiple R-squared: 0.276, Adjusted R-squared: 0.27
## F-statistic: 48.1 on 12 and 1515 DF, p-value: <2e-16#
lmn3 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours +SubEduOver +SubMismatch , data = p1)
summary(lmn3)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + SubEduOver + SubMismatch, data = p1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.0553 -0.2376 -0.0135 0.1915 1.6806
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.034604 0.073523 136.48 < 2e-16 ***
## Gender男 0.216976 0.019025 11.40 < 2e-16 ***
## J_year 0.023532 0.001699 13.85 < 2e-16 ***
## Sector國外學校 0.377727 0.071012 5.32 1.2e-07 ***
## Sector國立（公立） 0.159363 0.019150 8.32 < 2e-16 ***
## Region北北基 0.037633 0.051500 0.73 0.4651
## Region桃竹苗 0.161382 0.054782 2.95 0.0033 **
## Region中彰投 -0.053385 0.054275 -0.98 0.3255
## Region雲嘉南 -0.066893 0.053672 -1.25 0.2128
## Region高屏澎 -0.072617 0.053477 -1.36 0.1747
## Hours 0.004038 0.000958 4.22 2.6e-05 ***
## SubEduOver高於工作要求 -0.116271 0.024228 -4.80 1.8e-06 ***
## SubEduOver低於工作要求 -0.151612 0.031061 -4.88 1.2e-06 ***
## SubMismatch 0.053431 0.007851 6.81 1.4e-11 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.358 on 1514 degrees of freedom
## Multiple R-squared: 0.31, Adjusted R-squared: 0.305
## F-statistic: 52.4 on 13 and 1514 DF, p-value: <2e-16客觀過量教育一樣只有限縮的部份有顯著，客觀關聯效果蠻好的
自評過量、關聯效果一樣不錯……
#
p2 <- p[ p$EduLv %in% c("博士","碩士", "普通大學", "科技大學", "技術學院"), ]
#
lmn1 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + ObjOver , data = p2)
summary(lmn1)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + ObjOver, data = p2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9279 -0.2454 -0.0158 0.2084 1.8240
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.13273 0.07198 140.77 < 2e-16 ***
## Gender男 0.19143 0.01972 9.71 < 2e-16 ***
## J_year 0.02595 0.00192 13.49 < 2e-16 ***
## Sector國外學校 0.36819 0.07367 5.00 6.5e-07 ***
## Sector國立（公立） 0.18930 0.01998 9.48 < 2e-16 ***
## Region北北基 0.05165 0.05349 0.97 0.33442
## Region桃竹苗 0.16567 0.05690 2.91 0.00365 **
## Region中彰投 -0.03127 0.05652 -0.55 0.58011
## Region雲嘉南 -0.05496 0.05582 -0.98 0.32494
## Region高屏澎 -0.07999 0.05574 -1.44 0.15150
## Hours 0.00371 0.00100 3.70 0.00022 ***
## ObjOverover 0.01956 0.02103 0.93 0.35242
## ObjOverunder 0.14981 0.04963 3.02 0.00258 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.365 on 1448 degrees of freedom
## Multiple R-squared: 0.256, Adjusted R-squared: 0.25
## F-statistic: 41.5 on 12 and 1448 DF, p-value: <2e-16#
lmn2 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + Core , data = p2)
summary(lmn2)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + Core, data = p2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.9554 -0.2388 -0.0167 0.2015 1.8713
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.097824 0.070721 142.78 < 2e-16 ***
## Gender男 0.196441 0.019510 10.07 < 2e-16 ***
## J_year 0.026391 0.001905 13.85 < 2e-16 ***
## Sector國外學校 0.377379 0.072916 5.18 2.6e-07 ***
## Sector國立（公立） 0.180053 0.019632 9.17 < 2e-16 ***
## Region北北基 0.059829 0.052899 1.13 0.25824
## Region桃竹苗 0.175846 0.056170 3.13 0.00178 **
## Region中彰投 -0.013663 0.055843 -0.24 0.80675
## Region雲嘉南 -0.042197 0.055116 -0.77 0.44403
## Region高屏澎 -0.065873 0.055072 -1.20 0.23184
## Hours 0.003617 0.000992 3.65 0.00027 ***
## Core部分關聯 0.124463 0.028762 4.33 1.6e-05 ***
## Core核心關聯 0.125587 0.022369 5.61 2.4e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.361 on 1448 degrees of freedom
## Multiple R-squared: 0.272, Adjusted R-squared: 0.266
## F-statistic: 45.1 on 12 and 1448 DF, p-value: <2e-16#
lmn3 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours +SubEduOver +SubMismatch , data = p2)
summary(lmn3)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + SubEduOver + SubMismatch, data = p2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.061 -0.235 -0.009 0.193 1.674
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.03743 0.07398 135.68 < 2e-16 ***
## Gender男 0.20989 0.01920 10.93 < 2e-16 ***
## J_year 0.02445 0.00187 13.09 < 2e-16 ***
## Sector國外學校 0.38898 0.07153 5.44 6.3e-08 ***
## Sector國立（公立） 0.16748 0.01939 8.64 < 2e-16 ***
## Region北北基 0.03982 0.05181 0.77 0.4423
## Region桃竹苗 0.15738 0.05498 2.86 0.0043 **
## Region中彰投 -0.03097 0.05464 -0.57 0.5710
## Region雲嘉南 -0.05963 0.05393 -1.11 0.2691
## Region高屏澎 -0.07396 0.05390 -1.37 0.1702
## Hours 0.00399 0.00097 4.11 4.1e-05 ***
## SubEduOver高於工作要求 -0.11512 0.02435 -4.73 2.5e-06 ***
## SubEduOver低於工作要求 -0.14991 0.03188 -4.70 2.8e-06 ***
## SubMismatch 0.04998 0.00799 6.26 5.2e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.353 on 1447 degrees of freedom
## Multiple R-squared: 0.304, Adjusted R-squared: 0.298
## F-statistic: 48.6 on 13 and 1447 DF, p-value: <2e-16大學生們在客評過量、關聯的效果非常符合研究期待
在自評部份也表現得不錯~~
p3 <- p[ p$EduLv =="普通大學", ]
#
lmn1 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + ObjOver , data = p3)
summary(lmn1)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + ObjOver, data = p3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.8606 -0.2297 0.0024 0.1746 1.4490
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.39298 0.08886 116.96 < 2e-16 ***
## Gender男 0.09634 0.02620 3.68 0.00026 ***
## J_year 0.02696 0.00288 9.38 < 2e-16 ***
## Sector國外學校 0.29113 0.14654 1.99 0.04736 *
## Sector國立（公立） 0.05768 0.02500 2.31 0.02134 *
## Region北北基 -0.06319 0.06427 -0.98 0.32586
## Region桃竹苗 -0.03988 0.07091 -0.56 0.57404
## Region中彰投 -0.12833 0.06815 -1.88 0.06012 .
## Region雲嘉南 -0.20235 0.06901 -2.93 0.00348 **
## Region高屏澎 -0.14350 0.06785 -2.11 0.03481 *
## Hours 0.00267 0.00130 2.05 0.04065 *
## ObjOverover -0.07707 0.02554 -3.02 0.00265 **
## ObjOverunder 0.19290 0.05442 3.54 0.00042 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.321 on 676 degrees of freedom
## Multiple R-squared: 0.205, Adjusted R-squared: 0.191
## F-statistic: 14.5 on 12 and 676 DF, p-value: <2e-16#
lmn2 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours + Core , data = p3)
summary(lmn2)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + Core, data = p3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.8663 -0.2332 -0.0088 0.1737 1.5754
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.30022 0.08851 116.37 < 2e-16 ***
## Gender男 0.10452 0.02639 3.96 8.2e-05 ***
## J_year 0.02687 0.00290 9.25 < 2e-16 ***
## Sector國外學校 0.34506 0.14676 2.35 0.01900 *
## Sector國立（公立） 0.05909 0.02496 2.37 0.01818 *
## Region北北基 -0.04357 0.06471 -0.67 0.50099
## Region桃竹苗 -0.03278 0.07128 -0.46 0.64571
## Region中彰投 -0.10804 0.06852 -1.58 0.11532
## Region雲嘉南 -0.19559 0.06932 -2.82 0.00492 **
## Region高屏澎 -0.12056 0.06815 -1.77 0.07733 .
## Hours 0.00275 0.00131 2.10 0.03593 *
## Core部分關聯 0.13188 0.03908 3.37 0.00078 ***
## Core核心關聯 0.11616 0.02988 3.89 0.00011 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.322 on 676 degrees of freedom
## Multiple R-squared: 0.198, Adjusted R-squared: 0.183
## F-statistic: 13.9 on 12 and 676 DF, p-value: <2e-16#
lmn3 <- lm(log(Salary) ~ Gender + J_year + Sector +Region + Hours +SubEduOver +SubMismatch , data = p3)
summary(lmn3)##
## Call:
## lm(formula = log(Salary) ~ Gender + J_year + Sector + Region +
## Hours + SubEduOver + SubMismatch, data = p3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.7786 -0.2052 -0.0147 0.1705 1.5162
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.21838 0.09405 108.65 < 2e-16 ***
## Gender男 0.11493 0.02604 4.41 1.2e-05 ***
## J_year 0.02660 0.00285 9.34 < 2e-16 ***
## Sector國外學校 0.38019 0.14415 2.64 0.00855 **
## Sector國立（公立） 0.04566 0.02449 1.86 0.06269 .
## Region北北基 -0.05610 0.06351 -0.88 0.37739
## Region桃竹苗 -0.02489 0.07013 -0.35 0.72278
## Region中彰投 -0.11381 0.06732 -1.69 0.09135 .
## Region雲嘉南 -0.18531 0.06805 -2.72 0.00663 **
## Region高屏澎 -0.11480 0.06695 -1.71 0.08687 .
## Hours 0.00336 0.00128 2.62 0.00893 **
## SubEduOver高於工作要求 -0.11033 0.03337 -3.31 0.00099 ***
## SubEduOver低於工作要求 -0.11087 0.04288 -2.59 0.00993 **
## SubMismatch 0.04522 0.01034 4.37 1.4e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.316 on 675 degrees of freedom
## Multiple R-squared: 0.228, Adjusted R-squared: 0.214
## F-statistic: 15.4 on 13 and 675 DF, p-value: <2e-16