mobile

1 数据和变量

数据来源：mobile.xlsx

样本容量：142

laptop: 笔记本电脑价格(元)

cost: 月生活费(元)

mobile: 手机价格(元)

mobile: 笔记本电脑使用时长(月)

gender: 性别

demand_type: 需求类型(性能-performance/轻量-lite)

brand_prefer: 品牌偏好(高端设计-premium/主流商务-mainstream/专业游戏-gaming/高性价比-value)

2 描述性统计分析

2.1 定量变量

library(modelsummary)
datasummary(laptop + cost + mobile + usage_month ~ 
                     N + Mean + SD + Median + Min + Max, 
                   data = df, 
                   fmt = 3, 
                   output = 'flextable') %>%
  theme_apa() %>%              
  autofit() %>%               
  font(fontname = "Times New Roman", part = "all")

	N	Mean	SD	Median	Min	Max
laptop	142	6260.627	2362.171	6000.000	1800.000	13000.000
cost	142	2100.077	672.072	2000.000	800.000	5000.000
mobile	142	5329.824	2828.370	5299.500	1200.000	13999.000
usage_month	142	15.697	10.254	16.000	1.000	62.000

---

2.2 定性变量

datasummary(gender + demand_type + brand_prefer ~ 
            N + Percent() + laptop * (Mean + Median + SD + Min + Max), 
            data = df, 
            fmt = 2, 
            output = 'flextable') %>%
  theme_apa() %>%
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

		N	Percent	Mean	Median	SD	Min	Max
gender	female	108	76.06	5865.74	5550.00	2262.62	1800.00	13000.00
	male	34	23.94	7514.97	7149.50	2260.61	2599.00	12999.00
demand_type	lite	90	63.38	5733.38	5550.00	2120.69	1800.00	12499.00
	performance	52	36.62	7173.17	6894.50	2497.74	2300.00	13000.00
brand_prefer	gaming	21	14.79	8042.24	7800.00	1947.86	5400.00	13000.00
	mainstream	81	57.04	5800.21	5600.00	2078.70	1800.00	12999.00
	premium	18	12.68	7578.33	6857.50	2639.92	3599.00	12499.00
	value	22	15.49	5177.05	5000.00	2244.66	2000.00	10000.00

3 回归分析

3.1 Model 1

library(broom)
m1 <- lm(laptop ~ cost + mobile + usage_month, data = df)

models_list <- list("Model 1" = m1)

modelsummary(
  list("Model 1" = m1), 
  fmt = 3, 
  stars = TRUE, 
  gof_omit = 'AIC|BIC|Log.Lik|F|statistic', 
  output = "flextable"
) %>% 
  theme_apa() %>%
  fontsize(size = 18, part = "all") %>%      
  padding(padding = 5, part = "all") %>%
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

Model 1 Regression Analysis

	Model 1
(Intercept)	4689.049***
	(654.061)
cost	0.399
	(0.289)
mobile	0.291***
	(0.069)
usage_month	-52.193**
	(17.250)
Num.Obs.	142
R2	0.230
R2 Adj.	0.213
RMSE	2065.51
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001

\[ \operatorname{\widehat{laptop}} = 4689.049 + 0.399(\operatorname{cost}) + 0.291(\operatorname{mobile}) - 52.193(\operatorname{usage\_month}) \]

---

3.2 Model 2.1

library(broom)
m2 <- lm(log(laptop) ~ log(cost) + log(mobile) + usage_month, data = df)

models_list <- list("Model 2 (Y=log(laptop))" = m2)

modelsummary(
  models_list, 
  fmt = 3, 
  stars = TRUE,  
  gof_omit = 'AIC|BIC|Log.Lik|F|statistic', 
  output = "flextable"
) %>% 
  theme_apa() %>% 
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

Model 2 Regression Analysis

	Model 2 (Y=log(laptop))
(Intercept)	5.785***
	(0.735)
log(cost)	0.110
	(0.103)
log(mobile)	0.258***
	(0.056)
usage_month	-0.008**
	(0.003)
Num.Obs.	142
R2	0.229
R2 Adj.	0.213
RMSE	0.35
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001

\[ \operatorname{\widehat{log(laptop)}} = 5.785 + 0.11(\operatorname{\log(cost)}) + 0.258(\operatorname{\log(mobile)}) - 0.008(\operatorname{usage\_month}) \]

3.3 Model 2.2

library(broom)
m2.2 <- lm(log(laptop) ~ log(cost) + log(mobile) + usage_month + gender + 
     gender:log(mobile), data = df)

models_list <- list("Model 2 (Y=log(laptop))" = m2.2)

modelsummary(
  models_list, 
  fmt = 3, 
  stars = TRUE,  
  gof_omit = 'AIC|BIC|Log.Lik|F|statistic', 
  output = "flextable"
) %>% 
  theme_apa() %>% 
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

Model 2 Regression Analysis

	Model 2 (Y=log(laptop))
(Intercept)	5.027***
	(0.740)
log(cost)	0.122
	(0.096)
log(mobile)	0.325***
	(0.059)
usage_month	-0.006*
	(0.003)
gendermale	1.827*
	(0.920)
log(mobile) × gendermale	-0.183+
	(0.110)
Num.Obs.	142
R2	0.341
R2 Adj.	0.317
RMSE	0.32
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001

\[ \begin{aligned} \operatorname{\widehat{log(laptop)}} &= 5.027 + 0.122(\operatorname{\log(cost)}) + 0.325(\operatorname{\log(mobile)}) - 0.006(\operatorname{usage\_month})\ + \\ &\quad 1.827(\operatorname{gender}_{\operatorname{male}}) - 0.183(\operatorname{\log(mobile)} \times \operatorname{gender}_{\operatorname{male}}) \end{aligned} \]

---

3.4 Model 3

m3 <- lm(laptop ~ cost + mobile + usage_month + gender, data = df)

models_list <- list("Model 3" = m3)

modelsummary(
  models_list, 
  fmt = 3, 
  stars = TRUE,  
  gof_omit = 'AIC|BIC|Log.Lik|F|statistic', 
  output = "flextable"
) %>% 
  theme_apa() %>% 
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

Regression Analysis

	Model 3
(Intercept)	3905.511***
	(638.666)
cost	0.424
	(0.272)
mobile	0.317***
	(0.065)
usage_month	-40.910*
	(16.386)
gendermale	1745.878***
	(393.317)
Num.Obs.	142
R2	0.327
R2 Adj.	0.307
RMSE	1931.29
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001

\[ \operatorname{\widehat{laptop}} = 3905.511 + 0.424(\operatorname{cost}) + 0.317(\operatorname{mobile}) - 40.91(\operatorname{usage\_month}) + 1745.878(\operatorname{gender}_{\operatorname{male}}) \]

---

3.5 Model 4

m4 <- lm(log(laptop) ~ log(cost) + log(mobile) + 
           usage_month + 
           gender + demand_type + brand_prefer , data = df)

models_list <- list("Model 4" = m4)

modelsummary(
  models_list, 
  fmt = 3, 
  stars = TRUE,  
  gof_omit = 'AIC|BIC|Log.Lik|F|statistic', 
  output = "flextable"
) %>% 
  theme_apa() %>% 
  autofit() %>%
  font(fontname = "Times New Roman", part = "all")

Regression Analysis

	Model 4
(Intercept)	6.288***
	(0.730)
log(cost)	0.070
	(0.096)
log(mobile)	0.241***
	(0.052)
usage_month	-0.005*
	(0.003)
gendermale	0.231***
	(0.068)
demand_typeperformance	0.094
	(0.066)
brand_prefermainstream	-0.217*
	(0.087)
brand_preferpremium	-0.081
	(0.113)
brand_prefervalue	-0.322**
	(0.101)
Num.Obs.	142
R2	0.406
R2 Adj.	0.370
RMSE	0.31
+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001

\[ \begin{aligned} \operatorname{\widehat{log(laptop)}} &= 6.2883 + 0.07(\operatorname{\log(cost)}) + 0.2408(\operatorname{\log(mobile)}) - 0.0055(\operatorname{usage\_month})\ + \\ &\quad 0.2306(\operatorname{gender}_{\operatorname{male}}) + 0.094(\operatorname{demand\_type}_{\operatorname{performance}}) - 0.2174(\operatorname{brand\_prefer}_{\operatorname{mainstream}}) - 0.0812(\operatorname{brand\_prefer}_{\operatorname{premium}})\ - \\ &\quad 0.3221(\operatorname{brand\_prefer}_{\operatorname{value}}) \end{aligned} \]