Folder 'Data_and_Output' already exists. No action taken.
Pretest Descriptive Results
table(is.na(fullchoice$pref1)) ### should be all FALSE
FALSE
2140
table(fullchoice$block) ### should always be a multiple of 10 and ideally equal
1 2 3 4 5 6 7 8 9 10
220 210 210 220 210 210 220 220 210 210
table(fullchoice$arm)
1 2 3 4
530 540 530 540
<- fullchoice %>% group_by(RID) %>% summarize(n= n())
t
<- fullchoice_raw %>% group_by(RID) %>% summarize(n= n())
t_raw
table(t$n) ### should all be 10, how many choice sets per respondent
10
214
table(fullchoice$pref1) ## How many opt out choices
1 2
1160 980
table(data$exit_code) # must be zero
0
214
table(data$status_comp)
finished
214
table(data$status_comp2)
< table of extent 0 >
# Where do we "loose" respondents?
table(raw_data$q1) # Do you want to participate in the survey?
1 2
348 11
table(raw_data$status) # did api call started?
success
335
sum(!is.na(raw_data$q10)) # How many are still there when we ask first attention question?
[1] 315
sum(!is.na(raw_data$q14)) # How many are still there when we ask third attention question?
[1] 296
table(raw_data$status_comp) # Is R studio server calculation done?
finished
291
sum(!is.na(raw_data$q16)) # HNV map question
[1] 284
sum(!is.na(raw_data$q18)) # PA map question
[1] 282
table(t_raw$n) # We loose some respondents after the first choice card
1 2 3 4 5 6 7 8 9 10
43 4 2 1 2 1 4 1 2 219
<- t_raw %>% filter(n==1)
loose_first_cc <- unique(loose_first_cc$RID)
ids_first_cc
<- fullchoice_raw %>% filter(RID %in% ids_first_cc)
loose_first_cc
summary(loose_first_cc$pref1) # DCE display problem? Or do people just drop out?
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
NA NA NA NaN NA NA 43
table(as.factor(loose_first_cc$is_mobile))
true
43
sum(!is.na(raw_data$q27_1_1)) # First question after DCE
[1] 219
Meta data
Mobile device
Quality checks
Randomizations
Socio-demografics
# A tibble: 7 × 3
educ n percentage
<fct> <int> <dbl>
1 Hochschulabschluss (Universität, FH) 78 36.4
2 Mittlere Reife (Realschulabschluss) 53 24.8
3 Abitur 48 22.4
4 Hauptschulabschluss 28 13.1
5 Von der Schule abgegangen ohne Schulabschluss 3 1.40
6 Sonstiges 3 1.40
7 Will ich nicht beantworten 1 0.467
Attention questions
Access of Additional Information
1
40
1
34
NSG and HNV Distribution
Spatial Distribution
Choice Experiment
Conditional Logit Models
Estimate | s.e. | t.rat.(0) | Rob.s.e. | Rob.t.rat.(0) | |
beta_pa | 0.071 | 0.022 | 3.261 | 0.034 | 2.103 |
beta_hnv | 0.060 | 0.022 | 2.702 | 0.031 | 1.952 |
beta_pa_sq | -0.0001 | 0.0001 | -1.778 | 0.0002 | -0.923 |
beta_hnv_sq | -0.00004 | 0.0001 | -0.593 | 0.0001 | -0.319 |
pa_half_access | 0.002 | 0.109 | 0.020 | 0.103 | 0.021 |
pa_full_access | -0.046 | 0.103 | -0.444 | 0.103 | -0.444 |
hnv_visible | 0.035 | 0.090 | 0.392 | 0.080 | 0.439 |
beta_cost | -0.008 | 0.001 | -10.024 | 0.001 | -7.492 |
Estimate | s.e. | t.rat.(0) | Rob.s.e. | Rob.t.rat.(0) | |
beta_pa | 8.813 | 2.705 | 3.258 | 4.173 | 2.112 |
beta_hnv | 7.489 | 2.712 | 2.761 | 3.803 | 1.969 |
beta_pa_sq | -0.018 | 0.010 | -1.751 | 0.020 | -0.916 |
beta_hnv_sq | -0.005 | 0.008 | -0.588 | 0.016 | -0.317 |
pa_half_access | 0.310 | 13.570 | 0.023 | 12.879 | 0.024 |
pa_full_access | -5.579 | 13.006 | -0.429 | 13.035 | -0.428 |
hnv_visible | 4.501 | 11.215 | 0.401 | 10.044 | 0.448 |
beta_cost | -0.008 | 0.001 | -10.036 | 0.001 | -7.502 |
Estimate | s.e. | t.rat.(0) | Rob.s.e. | Rob.t.rat.(0) | |
beta_pa | 5.418 | 2.775 | 1.952 | 3.981 | 1.361 |
beta_hnv | 4.447 | 2.755 | 1.614 | 3.401 | 1.308 |
beta_pa_sq | -0.016 | 0.009 | -1.741 | 0.018 | -0.906 |
beta_hnv_sq | -0.005 | 0.008 | -0.659 | 0.014 | -0.353 |
pa_half_access | -8.918 | 12.983 | -0.687 | 11.930 | -0.748 |
pa_full_access | -18.082 | 13.079 | -1.383 | 12.678 | -1.426 |
hnv_visible | -2.218 | 10.616 | -0.209 | 9.305 | -0.238 |
beta_cost | -0.009 | 0.001 | -10.152 | 0.001 | -7.720 |
ASC_sq | -39.594 | 15.756 | -2.513 | 17.058 | -2.321 |
Estimate | s.e. | t.rat.(0) | Rob.s.e. | Rob.t.rat.(0) | |
beta_pa200 | 0.379 | 0.138 | 2.753 | 0.138 | 2.738 |
beta_pa300 | 0.355 | 0.145 | 2.444 | 0.122 | 2.911 |
beta_pa500 | 0.272 | 0.139 | 1.960 | 0.149 | 1.819 |
beta_pa800 | 0.425 | 0.129 | 3.299 | 0.119 | 3.577 |
beta_hnv200 | 0.063 | 0.137 | 0.461 | 0.147 | 0.430 |
beta_hnv300 | -0.0005 | 0.137 | -0.003 | 0.146 | -0.003 |
beta_hnv500 | -0.061 | 0.153 | -0.397 | 0.152 | -0.402 |
beta_hnv800 | 0.335 | 0.146 | 2.300 | 0.137 | 2.450 |
pa_half_access | 0.032 | 0.118 | 0.271 | 0.106 | 0.301 |
pa_full_access | -0.002 | 0.122 | -0.016 | 0.118 | -0.017 |
hnv_visible | 0.002 | 0.094 | 0.024 | 0.079 | 0.028 |
beta_cost | -0.008 | 0.001 | -9.836 | 0.001 | -7.361 |
Check Choice Patterns
# Relationship between Attribute Levels and Choices
table((database$pa_att - database$sq_pa_area), database$pref1)
1 2
100 259 170
200 231 196
300 216 211
500 231 197
800 223 206
prop.table(table((database$pa_att - database$sq_pa_area), database$pref1), margin = 1)
1 2
100 0.6037296 0.3962704
200 0.5409836 0.4590164
300 0.5058548 0.4941452
500 0.5397196 0.4602804
800 0.5198135 0.4801865
table(database$Dummy_pa_half, database$pref1)
1 2
0 789 646
1 371 334
prop.table(table(database$Dummy_pa_half, database$pref1), margin = 1)
1 2
0 0.5498258 0.4501742
1 0.5262411 0.4737589
table(database$Dummy_pa_full, database$pref1)
1 2
0 758 675
1 402 305
prop.table(table(database$Dummy_pa_full, database$pref1), margin = 1)
1 2
0 0.5289602 0.4710398
1 0.5685997 0.4314003
table((database$hnv_att - database$sq_hnv_area), database$pref1)
1 2
100 245 185
200 213 213
300 238 190
500 257 173
800 207 219
prop.table(table((database$hnv_att - database$sq_hnv_area), database$pref1), margin = 1)
1 2
100 0.5697674 0.4302326
200 0.5000000 0.5000000
300 0.5560748 0.4439252
500 0.5976744 0.4023256
800 0.4859155 0.5140845
table(database$Dummy_hnv_visible, database$pref1)
1 2
0 580 491
1 580 489
prop.table(table(database$Dummy_hnv_visible, database$pref1), margin = 1)
1 2
0 0.5415500 0.4584500
1 0.5425631 0.4574369
table(database$cost_att, database$pref1)
1 2
5 126 215
10 168 195
40 169 194
80 204 140
120 242 121
150 251 115
prop.table(table(database$cost_att, database$pref1), margin = 1)
1 2
5 0.3695015 0.6304985
10 0.4628099 0.5371901
40 0.4655647 0.5344353
80 0.5930233 0.4069767
120 0.6666667 0.3333333
150 0.6857923 0.3142077
table((database$hnv_att - database$sq_hnv_area + database$pa_att - database$sq_pa_area), database$pref1)
1 2
200 48 39
300 120 72
400 151 106
500 61 66
600 112 104
700 122 91
800 62 46
900 125 110
1000 109 103
1100 121 136
1300 121 93
1600 8 14
prop.table(table((database$hnv_att - database$sq_hnv_area + database$pa_att - database$sq_pa_area), database$pref1), margin=1)
1 2
200 0.5517241 0.4482759
300 0.6250000 0.3750000
400 0.5875486 0.4124514
500 0.4803150 0.5196850
600 0.5185185 0.4814815
700 0.5727700 0.4272300
800 0.5740741 0.4259259
900 0.5319149 0.4680851
1000 0.5141509 0.4858491
1100 0.4708171 0.5291829
1300 0.5654206 0.4345794
1600 0.3636364 0.6363636
# Has radius an impact?
table(database$radius, database$pref1)
1 2
15000 309 251
20000 254 286
25000 266 224
30000 331 219
prop.table(table(database$radius, database$pref1), margin = 1)
1 2
15000 0.5517857 0.4482143
20000 0.4703704 0.5296296
25000 0.5428571 0.4571429
30000 0.6018182 0.3981818
# Test impact of zooming
table(database$zoom_first_cc, database$pref1)
1 2
0 967 823
1 193 157
prop.table(table(database$zoom_first_cc, database$pref1), margin = 1)
1 2
0 0.5402235 0.4597765
1 0.5514286 0.4485714
# City vs Village people
prop.table(table(database$urban_rural, database$pref1), margin = 1)
1 2
Large City 0.4920000 0.5080000
Small City 0.5458824 0.4541176
Village 0.6076923 0.3923077
# Has DCE order an impact?
table(database$dce_version, database$pref1)
1 2
1 635 445
2 525 535
prop.table(table(database$dce_version, database$pref1), margin = 1)
1 2
1 0.587963 0.412037
2 0.495283 0.504717
# Check if it is a recoding issue
table(database$dce_version, database$pref1, database$cost_att)
, , = 5
1 2
1 72 102
2 54 113
, , = 10
1 2
1 102 80
2 66 115
, , = 40
1 2
1 86 93
2 83 101
, , = 80
1 2
1 107 67
2 97 73
, , = 120
1 2
1 129 55
2 113 66
, , = 150
1 2
1 139 48
2 112 67
# Has the order of attributes an impact?
table(database$order, database$pref1)
1 2
1,2,3,4,5 263 327
3,4,1,2,5 465 335
5,1,2,3,4 296 224
5,3,4,1,2 136 94
prop.table(table(database$order, database$pref1), margin = 1)
1 2
1,2,3,4,5 0.4457627 0.5542373
3,4,1,2,5 0.5812500 0.4187500
5,1,2,3,4 0.5692308 0.4307692
5,3,4,1,2 0.5913043 0.4086957
# Do correct responses have an impact?
table(database$corr_all, database$pref1)
1 2
0 918 762
1 242 218
prop.table(table(database$corr_all, database$pref1), margin = 1)
1 2
0 0.5464286 0.4535714
1 0.5260870 0.4739130
Protester
Follow-up Fragen
CV
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
0 0 20 2122 80 250000 82
Comments