Parenting Project Attitudes Questionnaire Experiment 6
Emily & Mike
2016-01-09
Data analysis of basic parenting values/attitudes survey, version 2.
1 Data preprocessing
Preliminaries.
## [1] "dplyr" "langcog" "tidyr" "ggplot2" "lme4"##
## Attaching package: 'langcog'
##
## The following object is masked from 'package:base':
##
## scale
##
##
## Attaching package: 'dplyr'
##
## The following objects are masked from 'package:plyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarize
##
## The following objects are masked from 'package:stats':
##
## filter, lag
##
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
##
##
## Attaching package: 'psych'
##
## The following object is masked from 'package:ggplot2':
##
## %+%
##
## Loading required package: MASS
##
## Attaching package: 'MASS'
##
## The following object is masked from 'package:dplyr':
##
## select
##
## Loading required package: boot
##
## Attaching package: 'boot'
##
## The following object is masked from 'package:psych':
##
## logit
##
## Loading required package: lattice
##
## Attaching package: 'lattice'
##
## The following object is masked from 'package:boot':
##
## melanoma
##
##
## Attaching package: 'nFactors'
##
## The following object is masked from 'package:lattice':
##
## parallelRead in files and consolidate to the same directory.
files <- dir("../production-results/e6/")
d.raw <- data.frame()
for (f in files) {
jf <- paste("../production-results/e6/",f,sep="")
jd <- fromJSON(paste(readLines(jf), collapse=""))
# clean up instruction trial
sent <- jd$answers$data$sentence
rating <- as.numeric(jd$answers$data$rating)
trial_type <- jd$answer$data$trial_type
sent <- sent[trial_type != "2afc_instructions"]
trial_type <- trial_type[trial_type != "2afc_instructions"]
#left out race because not sure yet how to deal with list format
id <- data.frame(workerid = jd$WorkerId,
sent = sent,
rating = rating,
children = jd$answers$data$children,
language = jd$answer$data$homelang,
ses = jd$answer$data$ladder,
gender = jd$answer$data$gender,
age = jd$answer$data$age,
education = jd$answer$data$education,
ethnicity = jd$answer$data$ethnicity,
childAgeYoung = jd$answer$data$childAgeYoung,
childAgeOld = jd$answer$data$childAgeOld
)
d.raw <- bind_rows(d.raw, id)
}Map on question short forms so that we can use these instead.
labels <- read.csv("sent_forms_e6.csv")
labels$sent <- as.character(labels$sent)Clean up labels.
d.raw$sent <- as.character(d.raw$sent)
d.raw$sent <- str_replace_all(d.raw$sent, "'", "")
d.raw$sent <- str_replace_all(d.raw$sent, "’", "")
d.raw$sent <- str_replace_all(d.raw$sent, "“", "")
d.raw$sent <- str_replace_all(d.raw$sent, "”", "")
d.raw$sent <- str_replace_all(d.raw$sent, "‘", "")
d.raw$sent <- str_replace_all(d.raw$sent, "â", "")Merge.
d <- left_join(d.raw, labels)
#set up reverse code variables so that items don't get reverse coded a second time.
d$reverse_code[d$reverse_code == 1 & d$rating == "0"] <- 0
d$reverse_code[d$reverse_code == 1 & d$rating == "1"] <- 1
d$reverse_code[d$reverse_code == 1 & d$rating == "2"] <- 2
d$reverse_code[d$reverse_code == 1 & d$rating == "4"] <- 4
d$reverse_code[d$reverse_code == 1 & d$rating == "5"] <- 5
d$reverse_code[d$reverse_code == 1 & d$rating == "6"] <- 6
#reverse code.
d$rating[d$reverse_code == 0 & d$rating == "0"] <- 6
d$rating[d$reverse_code == 1 & d$rating == "1"] <- 5
d$rating[d$reverse_code == 2 & d$rating == "2"] <- 4
d$rating[d$reverse_code == 4 & d$rating == "4"] <- 2
d$rating[d$reverse_code == 5 & d$rating == "5"] <- 1
d$rating[d$reverse_code == 6 & d$rating == "6"] <- 0Plot demographic info.
subinfo <- d %>%
group_by(workerid) %>%
summarise(age = age[1],
gender = gender[1],
children = children[1],
ses = ses[1],
education = education[1],
language = language[1],
ethnicity = ethnicity[1],
youngestChildAge = childAgeYoung[1],
oldestChildAge = childAgeOld[1]
)
subinfo$education <- factor(subinfo$education, levels = c("highSchool","someCollege","4year","someGrad","Grad"))
subinfo$gender <- str_replace_all(subinfo$gender, "female|FEMALE|F$|f$|Femal$|Females|Females","Female")
subinfo$gender <- str_replace_all(subinfo$gender, "^male|^Male|^MALE|^M$|^m$|^Maleq|Make", "Male")
subinfo$gender <- str_replace_all(subinfo$gender, "29|24|25|28|32|33|45", "")
subinfo$gender <- str_replace_all(subinfo$gender, "males", "male")
subinfo$gender <- str_replace_all(subinfo$gender, " ", "")
subinfo$language <- str_replace_all(subinfo$language, "english|eNGLISH|Engliah|ENGLISH|^eng$|Enlgish", "English")
subinfo$language <- str_replace_all(subinfo$language," ", "")
subinfo$language <- str_replace_all(subinfo$language,"arabic", "Arabic")
subinfo$language <- str_replace_all(subinfo$language,"chinese", "Chinese")
subinfo$language <- str_replace_all(subinfo$language,"german", "German")
subinfo$language <- str_replace_all(subinfo$language,"tagalog", "Tagalog")
subinfo$youngestChildAge <- factor(subinfo$youngestChildAge, levels = c("","0to6mo","7to12mo","1y","2y","3y","4y","5y","6y","7y","8y","9y","10y","olderthan10"))
subinfo$oldestChildAge <- factor(subinfo$oldestChildAge, levels = c("","0to6mo","7to12mo","1y","2y","3y","4y","5y","6y","7y","8y","9y","10y","olderthan10"))
qplot(ses, data=subinfo)
qplot(children, data=subinfo)
qplot(gender, data=subinfo)
qplot(education, data=subinfo)
qplot(age, data=subinfo)
qplot(language, data=subinfo)
qplot(ethnicity, data=subinfo)
qplot(youngestChildAge, data=subinfo)
qplot(oldestChildAge, data=subinfo)
2 Basic analyses
Now look at mean ratings across sentences.
rating_count <- table(d$rating)
rating_count##
## 0 1 2 3 4 5 6
## 204 452 613 1406 1705 2155 3965prop.table(rating_count) ##
## 0 1 2 3 4 5
## 0.01942857 0.04304762 0.05838095 0.13390476 0.16238095 0.20523810
## 6
## 0.37761905ms <- d %>%
group_by(category, instrument, short_sent) %>%
multi_boot_standard(col = "rating") %>%
arrange(instrument, category, desc(mean))
ms$short_sent_ord <- factor(ms$short_sent,
levels = ms$short_sent)Plot attitude.
Ratings are clearly lower for the reverse coded items (the final 2 items in each category); this could be due to response sets (i.e., participants endorsing all items highly).
qplot(short_sent_ord, mean, col = category,
ymin = ci_lower, ymax = ci_upper,
geom = "pointrange",
data = filter(ms, instrument == "attitudes")) +
theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = .5)) +
xlab("") +
ylab("Mean Rating") +
ylim(c(0,6)) +
scale_colour_solarized()
3 Scale reliability
3.1 Whole Scale
This analysis is not especially meaningful as it averages across the three subscales. Raw alpha is .88 which is higher than previous runs.
wide.attitudes <- d %>%
filter(instrument == "attitudes") %>%
select(workerid, short_sent, rating) %>%
spread(short_sent, rating)
alpha.mat <- as.matrix(select(wide.attitudes, -workerid))
alpha(x = alpha.mat)## Some items ( rules prevent tantrums ) were negatively correlated with the total scale and probably should be reversed. To do this, run the function again with the 'check.keys=TRUE' option##
## Reliability analysis
## Call: alpha(x = alpha.mat)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd
## 0.89 0.91 0.94 0.19 10 0.012 4.5 0.6
##
## lower alpha upper 95% confidence boundaries
## 0.87 0.89 0.92
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r
## can learn good and bad 0.89 0.90 0.94 0.19
## children should be grateful 0.89 0.91 0.94 0.19
## control behavior 0.89 0.91 0.94 0.20
## control emotions if comforted 0.89 0.91 0.94 0.20
## dont impose structure 0.89 0.91 0.94 0.20
## dont interrupt 0.89 0.91 0.94 0.20
## examples are good 0.89 0.91 0.94 0.19
## follow-in good 0.89 0.91 0.94 0.20
## give children comfort 0.89 0.91 0.94 0.19
## holding and cradling 0.89 0.90 0.94 0.19
## kids should not decide 0.89 0.91 0.94 0.20
## learn before talking 0.89 0.90 0.94 0.19
## learn by playing 0.89 0.90 0.94 0.19
## learn contingencies 0.89 0.91 0.94 0.19
## learn from other children 0.89 0.91 0.94 0.19
## learn from repetetive behaviors 0.89 0.91 0.94 0.19
## learn math before school 0.89 0.90 0.94 0.19
## new experiences 0.89 0.91 0.94 0.20
## nonsense games 0.89 0.91 0.94 0.19
## not spoiled with too much attention 0.89 0.91 0.94 0.19
## parents choose toys 0.89 0.91 0.94 0.19
## play freely 0.89 0.91 0.94 0.19
## play pretend 0.89 0.91 0.94 0.19
## play with friends 0.89 0.91 0.94 0.19
## punish for small rules 0.89 0.91 0.94 0.20
## punish misbehavior 0.89 0.91 0.94 0.20
## reading before school 0.89 0.91 0.94 0.19
## reading before speaking 0.89 0.91 0.94 0.19
## reasons for rules 0.89 0.91 0.94 0.19
## respect parents teachers 0.89 0.91 0.94 0.19
## rules prevent tantrums 0.90 0.91 0.94 0.20
## safe loving environment 0.89 0.90 0.93 0.19
## simple toys 0.89 0.91 0.94 0.20
## spontaneous play 0.89 0.91 0.94 0.20
## strict rules 0.89 0.91 0.94 0.19
## strong bond mom 0.89 0.90 0.94 0.19
## talk about emotions 0.89 0.90 0.93 0.19
## talk about feelings before mature 0.89 0.91 0.94 0.19
## talk about opinions 0.89 0.91 0.94 0.19
## too much affection does not harm 0.89 0.91 0.94 0.19
## wait when told 0.89 0.91 0.94 0.19
## worry about misbehavior 0.89 0.91 0.94 0.19
## S/N alpha se
## can learn good and bad 9.5 0.012
## children should be grateful 9.9 0.012
## control behavior 10.0 0.012
## control emotions if comforted 10.2 0.012
## dont impose structure 9.9 0.012
## dont interrupt 10.0 0.012
## examples are good 9.6 0.012
## follow-in good 10.0 0.012
## give children comfort 9.7 0.012
## holding and cradling 9.4 0.012
## kids should not decide 10.0 0.012
## learn before talking 9.5 0.012
## learn by playing 9.5 0.012
## learn contingencies 9.7 0.012
## learn from other children 9.9 0.012
## learn from repetetive behaviors 9.8 0.012
## learn math before school 9.5 0.012
## new experiences 10.3 0.012
## nonsense games 9.6 0.012
## not spoiled with too much attention 9.9 0.012
## parents choose toys 9.9 0.012
## play freely 9.8 0.012
## play pretend 9.5 0.012
## play with friends 9.7 0.012
## punish for small rules 10.1 0.012
## punish misbehavior 10.3 0.012
## reading before school 9.6 0.012
## reading before speaking 9.6 0.012
## reasons for rules 9.7 0.012
## respect parents teachers 9.6 0.012
## rules prevent tantrums 10.6 0.011
## safe loving environment 9.4 0.012
## simple toys 10.1 0.012
## spontaneous play 10.0 0.012
## strict rules 9.9 0.012
## strong bond mom 9.5 0.012
## talk about emotions 9.4 0.013
## talk about feelings before mature 9.6 0.012
## talk about opinions 9.6 0.012
## too much affection does not harm 9.7 0.012
## wait when told 9.7 0.012
## worry about misbehavior 9.9 0.012
##
## Item statistics
## n raw.r std.r r.cor r.drop mean
## can learn good and bad 250 0.628 0.660 0.657 0.604 5.3
## children should be grateful 250 0.394 0.407 0.394 0.344 4.5
## control behavior 250 0.292 0.313 0.293 0.241 4.4
## control emotions if comforted 250 0.242 0.205 0.177 0.176 3.1
## dont impose structure 250 0.371 0.349 0.333 0.312 3.9
## dont interrupt 250 0.320 0.320 0.304 0.269 4.2
## examples are good 250 0.554 0.582 0.574 0.523 5.1
## follow-in good 250 0.295 0.316 0.289 0.246 4.3
## give children comfort 250 0.508 0.526 0.512 0.475 5.1
## holding and cradling 250 0.672 0.689 0.688 0.646 5.2
## kids should not decide 250 0.356 0.334 0.313 0.300 3.9
## learn before talking 250 0.604 0.636 0.631 0.577 5.3
## learn by playing 250 0.627 0.647 0.642 0.602 5.3
## learn contingencies 250 0.510 0.537 0.528 0.477 5.0
## learn from other children 250 0.438 0.414 0.395 0.384 4.1
## learn from repetetive behaviors 250 0.466 0.477 0.462 0.423 4.6
## learn math before school 250 0.651 0.662 0.661 0.622 5.3
## new experiences 250 0.153 0.135 0.097 0.089 2.9
## nonsense games 250 0.584 0.586 0.577 0.549 4.8
## not spoiled with too much attention 250 0.415 0.381 0.369 0.350 3.5
## parents choose toys 250 0.387 0.364 0.344 0.327 3.9
## play freely 250 0.473 0.456 0.443 0.423 3.9
## play pretend 250 0.620 0.626 0.620 0.588 5.0
## play with friends 250 0.500 0.489 0.472 0.450 4.7
## punish for small rules 250 0.273 0.263 0.239 0.217 3.7
## punish misbehavior 250 0.147 0.132 0.090 0.081 3.4
## reading before school 250 0.594 0.605 0.603 0.560 5.3
## reading before speaking 250 0.604 0.605 0.598 0.568 5.3
## reasons for rules 250 0.546 0.534 0.521 0.501 4.8
## respect parents teachers 250 0.539 0.552 0.546 0.505 5.0
## rules prevent tantrums 250 -0.031 -0.057 -0.099 -0.102 2.4
## safe loving environment 250 0.682 0.712 0.715 0.665 5.6
## simple toys 250 0.239 0.247 0.218 0.181 3.9
## spontaneous play 250 0.330 0.340 0.314 0.282 4.4
## strict rules 250 0.416 0.391 0.366 0.359 3.9
## strong bond mom 250 0.610 0.638 0.635 0.586 5.4
## talk about emotions 250 0.690 0.705 0.707 0.663 5.3
## talk about feelings before mature 250 0.612 0.609 0.604 0.577 5.0
## talk about opinions 250 0.573 0.595 0.588 0.546 5.1
## too much affection does not harm 250 0.513 0.486 0.480 0.456 4.0
## wait when told 250 0.464 0.484 0.473 0.425 4.8
## worry about misbehavior 250 0.391 0.382 0.361 0.338 4.3
## sd
## can learn good and bad 0.98
## children should be grateful 1.45
## control behavior 1.37
## control emotions if comforted 1.73
## dont impose structure 1.65
## dont interrupt 1.40
## examples are good 1.08
## follow-in good 1.30
## give children comfort 1.06
## holding and cradling 1.14
## kids should not decide 1.56
## learn before talking 1.02
## learn by playing 1.03
## learn contingencies 1.07
## learn from other children 1.63
## learn from repetetive behaviors 1.35
## learn math before school 1.19
## new experiences 1.62
## nonsense games 1.29
## not spoiled with too much attention 1.89
## parents choose toys 1.70
## play freely 1.54
## play pretend 1.23
## play with friends 1.61
## punish for small rules 1.50
## punish misbehavior 1.67
## reading before school 1.26
## reading before speaking 1.39
## reasons for rules 1.54
## respect parents teachers 1.19
## rules prevent tantrums 1.79
## safe loving environment 0.78
## simple toys 1.52
## spontaneous play 1.31
## strict rules 1.68
## strong bond mom 0.94
## talk about emotions 1.22
## talk about feelings before mature 1.35
## talk about opinions 1.00
## too much affection does not harm 1.86
## wait when told 1.19
## worry about misbehavior 1.51
##
## Non missing response frequency for each item
## 0 1 2 3 4 5 6
## can learn good and bad 0.00 0.00 0.01 0.06 0.09 0.28 0.56
## children should be grateful 0.00 0.03 0.06 0.20 0.16 0.18 0.37
## control behavior 0.00 0.03 0.05 0.20 0.20 0.24 0.27
## control emotions if comforted 0.08 0.13 0.15 0.24 0.18 0.11 0.11
## dont impose structure 0.03 0.07 0.10 0.17 0.23 0.18 0.22
## dont interrupt 0.00 0.04 0.06 0.24 0.21 0.21 0.24
## examples are good 0.00 0.01 0.01 0.08 0.13 0.29 0.48
## follow-in good 0.00 0.03 0.06 0.18 0.24 0.28 0.20
## give children comfort 0.00 0.01 0.02 0.06 0.13 0.30 0.48
## holding and cradling 0.00 0.01 0.02 0.07 0.12 0.22 0.56
## kids should not decide 0.03 0.07 0.07 0.21 0.24 0.21 0.16
## learn before talking 0.00 0.01 0.01 0.04 0.12 0.20 0.62
## learn by playing 0.00 0.00 0.02 0.06 0.13 0.22 0.57
## learn contingencies 0.00 0.00 0.02 0.08 0.16 0.32 0.41
## learn from other children 0.04 0.04 0.07 0.17 0.24 0.18 0.27
## learn from repetetive behaviors 0.00 0.02 0.06 0.16 0.14 0.27 0.35
## learn math before school 0.00 0.01 0.03 0.05 0.09 0.16 0.66
## new experiences 0.08 0.15 0.16 0.29 0.16 0.08 0.08
## nonsense games 0.01 0.02 0.03 0.08 0.21 0.27 0.38
## not spoiled with too much attention 0.05 0.15 0.13 0.18 0.15 0.12 0.22
## parents choose toys 0.03 0.08 0.10 0.18 0.18 0.21 0.22
## play freely 0.02 0.05 0.09 0.23 0.23 0.18 0.20
## play pretend 0.00 0.01 0.03 0.11 0.16 0.22 0.48
## play with friends 0.04 0.03 0.03 0.07 0.15 0.25 0.43
## punish for small rules 0.05 0.05 0.08 0.23 0.29 0.20 0.10
## punish misbehavior 0.00 0.14 0.20 0.26 0.12 0.11 0.18
## reading before school 0.01 0.02 0.02 0.03 0.08 0.14 0.69
## reading before speaking 0.02 0.02 0.03 0.04 0.04 0.13 0.72
## reasons for rules 0.02 0.03 0.04 0.08 0.16 0.20 0.46
## respect parents teachers 0.00 0.02 0.02 0.07 0.19 0.24 0.46
## rules prevent tantrums 0.20 0.17 0.15 0.20 0.16 0.06 0.06
## safe loving environment 0.00 0.00 0.01 0.03 0.06 0.12 0.78
## simple toys 0.00 0.07 0.10 0.26 0.20 0.17 0.20
## spontaneous play 0.00 0.02 0.06 0.21 0.16 0.30 0.25
## strict rules 0.04 0.08 0.08 0.15 0.27 0.18 0.20
## strong bond mom 0.00 0.00 0.01 0.06 0.08 0.22 0.63
## talk about emotions 0.01 0.02 0.03 0.03 0.10 0.20 0.62
## talk about feelings before mature 0.01 0.03 0.03 0.06 0.12 0.23 0.52
## talk about opinions 0.00 0.00 0.01 0.08 0.14 0.29 0.48
## too much affection does not harm 0.04 0.09 0.07 0.16 0.18 0.12 0.34
## wait when told 0.00 0.01 0.03 0.11 0.17 0.31 0.37
## worry about misbehavior 0.01 0.04 0.08 0.16 0.24 0.21 0.27
## miss
## can learn good and bad 0
## children should be grateful 0
## control behavior 0
## control emotions if comforted 0
## dont impose structure 0
## dont interrupt 0
## examples are good 0
## follow-in good 0
## give children comfort 0
## holding and cradling 0
## kids should not decide 0
## learn before talking 0
## learn by playing 0
## learn contingencies 0
## learn from other children 0
## learn from repetetive behaviors 0
## learn math before school 0
## new experiences 0
## nonsense games 0
## not spoiled with too much attention 0
## parents choose toys 0
## play freely 0
## play pretend 0
## play with friends 0
## punish for small rules 0
## punish misbehavior 0
## reading before school 0
## reading before speaking 0
## reasons for rules 0
## respect parents teachers 0
## rules prevent tantrums 0
## safe loving environment 0
## simple toys 0
## spontaneous play 0
## strict rules 0
## strong bond mom 0
## talk about emotions 0
## talk about feelings before mature 0
## talk about opinions 0
## too much affection does not harm 0
## wait when told 0
## worry about misbehavior 03.2 Rules and Respect Subscale
Raw alpha is .74 here.
wide.rules_respect <- d %>%
filter(category == "rules_respect") %>%
select(workerid, short_sent, rating) %>%
spread(short_sent, rating)
alpha.rr <- as.matrix(select(wide.rules_respect, -workerid))
alpha(x = alpha.rr)##
## Reliability analysis
## Call: alpha(x = alpha.rr)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd
## 0.74 0.76 0.78 0.2 3.1 0.032 4 0.76
##
## lower alpha upper 95% confidence boundaries
## 0.67 0.74 0.8
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N
## children should be grateful 0.71 0.73 0.75 0.20 2.7
## control behavior 0.71 0.73 0.76 0.20 2.8
## dont impose structure 0.71 0.73 0.75 0.20 2.8
## dont interrupt 0.71 0.73 0.75 0.19 2.6
## kids should not decide 0.72 0.74 0.76 0.21 2.8
## punish for small rules 0.71 0.73 0.76 0.20 2.8
## punish misbehavior 0.75 0.77 0.79 0.23 3.3
## respect parents teachers 0.71 0.72 0.75 0.19 2.6
## rules prevent tantrums 0.76 0.77 0.79 0.24 3.4
## strict rules 0.72 0.74 0.77 0.21 2.8
## wait when told 0.71 0.72 0.75 0.19 2.6
## worry about misbehavior 0.71 0.73 0.76 0.20 2.7
## alpha se
## children should be grateful 0.035
## control behavior 0.034
## dont impose structure 0.035
## dont interrupt 0.035
## kids should not decide 0.034
## punish for small rules 0.035
## punish misbehavior 0.031
## respect parents teachers 0.035
## rules prevent tantrums 0.030
## strict rules 0.034
## wait when told 0.035
## worry about misbehavior 0.035
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## children should be grateful 250 0.54 0.57 0.54 0.418 4.5 1.5
## control behavior 250 0.53 0.55 0.50 0.414 4.4 1.4
## dont impose structure 250 0.57 0.55 0.51 0.430 3.9 1.7
## dont interrupt 250 0.60 0.62 0.59 0.484 4.2 1.4
## kids should not decide 250 0.53 0.51 0.46 0.391 3.9 1.6
## punish for small rules 250 0.55 0.55 0.49 0.424 3.7 1.5
## punish misbehavior 250 0.29 0.28 0.15 0.116 3.4 1.7
## respect parents teachers 250 0.59 0.63 0.61 0.500 5.0 1.2
## rules prevent tantrums 250 0.28 0.24 0.10 0.086 2.4 1.8
## strict rules 250 0.53 0.51 0.43 0.377 3.9 1.7
## wait when told 250 0.60 0.64 0.61 0.512 4.8 1.2
## worry about misbehavior 250 0.60 0.59 0.54 0.474 4.3 1.5
##
## Non missing response frequency for each item
## 0 1 2 3 4 5 6 miss
## children should be grateful 0.00 0.03 0.06 0.20 0.16 0.18 0.37 0
## control behavior 0.00 0.03 0.05 0.20 0.20 0.24 0.27 0
## dont impose structure 0.03 0.07 0.10 0.17 0.23 0.18 0.22 0
## dont interrupt 0.00 0.04 0.06 0.24 0.21 0.21 0.24 0
## kids should not decide 0.03 0.07 0.07 0.21 0.24 0.21 0.16 0
## punish for small rules 0.05 0.05 0.08 0.23 0.29 0.20 0.10 0
## punish misbehavior 0.00 0.14 0.20 0.26 0.12 0.11 0.18 0
## respect parents teachers 0.00 0.02 0.02 0.07 0.19 0.24 0.46 0
## rules prevent tantrums 0.20 0.17 0.15 0.20 0.16 0.06 0.06 0
## strict rules 0.04 0.08 0.08 0.15 0.27 0.18 0.20 0
## wait when told 0.00 0.01 0.03 0.11 0.17 0.31 0.37 0
## worry about misbehavior 0.01 0.04 0.08 0.16 0.24 0.21 0.27 03.3 Affection and Attachment Subscale
Raw alpha is .81, which is up from .57 previously!
wide.affection <- d %>%
filter(category == "affection") %>%
select(workerid, short_sent, rating) %>%
spread(short_sent, rating)
alpha.af <- as.matrix(select(wide.affection, -workerid))
alpha(x = alpha.af)##
## Reliability analysis
## Call: alpha(x = alpha.af)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd
## 0.81 0.83 0.85 0.34 5 0.027 4.7 0.82
##
## lower alpha upper 95% confidence boundaries
## 0.75 0.81 0.86
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r
## control emotions if comforted 0.82 0.84 0.85 0.37
## give children comfort 0.79 0.82 0.83 0.33
## holding and cradling 0.78 0.81 0.82 0.32
## not spoiled with too much attention 0.79 0.83 0.83 0.34
## safe loving environment 0.79 0.81 0.83 0.32
## strong bond mom 0.79 0.81 0.83 0.33
## talk about emotions 0.79 0.82 0.83 0.33
## talk about feelings before mature 0.79 0.82 0.83 0.34
## talk about opinions 0.79 0.82 0.84 0.34
## too much affection does not harm 0.77 0.81 0.82 0.33
## S/N alpha se
## control emotions if comforted 5.4 0.028
## give children comfort 4.5 0.030
## holding and cradling 4.2 0.031
## not spoiled with too much attention 4.7 0.030
## safe loving environment 4.3 0.030
## strong bond mom 4.3 0.030
## talk about emotions 4.4 0.030
## talk about feelings before mature 4.6 0.030
## talk about opinions 4.6 0.030
## too much affection does not harm 4.4 0.032
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## control emotions if comforted 250 0.51 0.42 0.32 0.32 3.1 1.73
## give children comfort 250 0.61 0.64 0.59 0.51 5.1 1.06
## holding and cradling 250 0.70 0.74 0.71 0.61 5.2 1.14
## not spoiled with too much attention 250 0.67 0.58 0.53 0.51 3.5 1.89
## safe loving environment 250 0.63 0.70 0.67 0.56 5.6 0.78
## strong bond mom 250 0.62 0.69 0.65 0.55 5.4 0.94
## talk about emotions 250 0.61 0.66 0.62 0.50 5.3 1.22
## talk about feelings before mature 250 0.60 0.63 0.57 0.48 5.0 1.35
## talk about opinions 250 0.56 0.61 0.55 0.46 5.1 1.00
## too much affection does not harm 250 0.75 0.68 0.65 0.62 4.0 1.86
##
## Non missing response frequency for each item
## 0 1 2 3 4 5 6
## control emotions if comforted 0.08 0.13 0.15 0.24 0.18 0.11 0.11
## give children comfort 0.00 0.01 0.02 0.06 0.13 0.30 0.48
## holding and cradling 0.00 0.01 0.02 0.07 0.12 0.22 0.56
## not spoiled with too much attention 0.05 0.15 0.13 0.18 0.15 0.12 0.22
## safe loving environment 0.00 0.00 0.01 0.03 0.06 0.12 0.78
## strong bond mom 0.00 0.00 0.01 0.06 0.08 0.22 0.63
## talk about emotions 0.01 0.02 0.03 0.03 0.10 0.20 0.62
## talk about feelings before mature 0.01 0.03 0.03 0.06 0.12 0.23 0.52
## talk about opinions 0.00 0.00 0.01 0.08 0.14 0.29 0.48
## too much affection does not harm 0.04 0.09 0.07 0.16 0.18 0.12 0.34
## miss
## control emotions if comforted 0
## give children comfort 0
## holding and cradling 0
## not spoiled with too much attention 0
## safe loving environment 0
## strong bond mom 0
## talk about emotions 0
## talk about feelings before mature 0
## talk about opinions 0
## too much affection does not harm 03.4 Active Learning Subscale
Raw alpha is .86, which is much higher than previously!
wide.active_learning <- d %>%
filter(category == "active_learning") %>%
select(workerid, short_sent, rating) %>%
spread(short_sent, rating)
alpha.al <- as.matrix(select(wide.active_learning, -workerid))
alpha(x = alpha.al)##
## Reliability analysis
## Call: alpha(x = alpha.al)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd
## 0.86 0.87 0.89 0.25 6.7 0.017 4.7 0.7
##
## lower alpha upper 95% confidence boundaries
## 0.82 0.86 0.89
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N
## can learn good and bad 0.85 0.86 0.88 0.25 6.2
## examples are good 0.85 0.86 0.88 0.25 6.4
## follow-in good 0.86 0.87 0.89 0.26 6.8
## learn before talking 0.85 0.86 0.88 0.25 6.2
## learn by playing 0.85 0.86 0.88 0.24 6.1
## learn contingencies 0.85 0.86 0.88 0.25 6.4
## learn from other children 0.85 0.87 0.88 0.26 6.5
## learn from repetetive behaviors 0.85 0.86 0.88 0.25 6.3
## learn math before school 0.85 0.86 0.88 0.24 6.1
## new experiences 0.87 0.88 0.89 0.27 7.1
## nonsense games 0.85 0.86 0.88 0.25 6.2
## parents choose toys 0.85 0.87 0.88 0.26 6.5
## play freely 0.85 0.87 0.88 0.25 6.4
## play pretend 0.85 0.86 0.88 0.25 6.2
## play with friends 0.85 0.86 0.88 0.25 6.3
## reading before school 0.85 0.86 0.88 0.25 6.2
## reading before speaking 0.85 0.86 0.88 0.25 6.3
## reasons for rules 0.85 0.86 0.88 0.25 6.4
## simple toys 0.86 0.87 0.89 0.27 6.9
## spontaneous play 0.86 0.87 0.89 0.26 6.6
## alpha se
## can learn good and bad 0.018
## examples are good 0.018
## follow-in good 0.018
## learn before talking 0.019
## learn by playing 0.019
## learn contingencies 0.018
## learn from other children 0.018
## learn from repetetive behaviors 0.018
## learn math before school 0.019
## new experiences 0.017
## nonsense games 0.019
## parents choose toys 0.018
## play freely 0.018
## play pretend 0.019
## play with friends 0.019
## reading before school 0.019
## reading before speaking 0.018
## reasons for rules 0.018
## simple toys 0.017
## spontaneous play 0.018
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## can learn good and bad 250 0.60 0.64 0.62 0.55 5.3 0.98
## examples are good 250 0.51 0.55 0.52 0.45 5.1 1.08
## follow-in good 250 0.36 0.37 0.31 0.27 4.3 1.30
## learn before talking 250 0.62 0.65 0.64 0.57 5.3 1.02
## learn by playing 250 0.65 0.67 0.66 0.60 5.3 1.03
## learn contingencies 250 0.52 0.55 0.52 0.46 5.0 1.07
## learn from other children 250 0.51 0.48 0.44 0.42 4.1 1.63
## learn from repetetive behaviors 250 0.56 0.56 0.53 0.49 4.6 1.35
## learn math before school 250 0.65 0.66 0.65 0.60 5.3 1.19
## new experiences 250 0.27 0.24 0.17 0.16 2.9 1.62
## nonsense games 250 0.64 0.64 0.62 0.59 4.8 1.29
## parents choose toys 250 0.52 0.49 0.45 0.43 3.9 1.70
## play freely 250 0.55 0.53 0.49 0.47 3.9 1.54
## play pretend 250 0.64 0.65 0.63 0.59 5.0 1.23
## play with friends 250 0.60 0.57 0.54 0.51 4.7 1.61
## reading before school 250 0.62 0.63 0.62 0.56 5.3 1.26
## reading before speaking 250 0.58 0.59 0.56 0.51 5.3 1.39
## reasons for rules 250 0.58 0.56 0.53 0.50 4.8 1.54
## simple toys 250 0.29 0.29 0.23 0.19 3.9 1.52
## spontaneous play 250 0.46 0.45 0.40 0.38 4.4 1.31
##
## Non missing response frequency for each item
## 0 1 2 3 4 5 6 miss
## can learn good and bad 0.00 0.00 0.01 0.06 0.09 0.28 0.56 0
## examples are good 0.00 0.01 0.01 0.08 0.13 0.29 0.48 0
## follow-in good 0.00 0.03 0.06 0.18 0.24 0.28 0.20 0
## learn before talking 0.00 0.01 0.01 0.04 0.12 0.20 0.62 0
## learn by playing 0.00 0.00 0.02 0.06 0.13 0.22 0.57 0
## learn contingencies 0.00 0.00 0.02 0.08 0.16 0.32 0.41 0
## learn from other children 0.04 0.04 0.07 0.17 0.24 0.18 0.27 0
## learn from repetetive behaviors 0.00 0.02 0.06 0.16 0.14 0.27 0.35 0
## learn math before school 0.00 0.01 0.03 0.05 0.09 0.16 0.66 0
## new experiences 0.08 0.15 0.16 0.29 0.16 0.08 0.08 0
## nonsense games 0.01 0.02 0.03 0.08 0.21 0.27 0.38 0
## parents choose toys 0.03 0.08 0.10 0.18 0.18 0.21 0.22 0
## play freely 0.02 0.05 0.09 0.23 0.23 0.18 0.20 0
## play pretend 0.00 0.01 0.03 0.11 0.16 0.22 0.48 0
## play with friends 0.04 0.03 0.03 0.07 0.15 0.25 0.43 0
## reading before school 0.01 0.02 0.02 0.03 0.08 0.14 0.69 0
## reading before speaking 0.02 0.02 0.03 0.04 0.04 0.13 0.72 0
## reasons for rules 0.02 0.03 0.04 0.08 0.16 0.20 0.46 0
## simple toys 0.00 0.07 0.10 0.26 0.20 0.17 0.20 0
## spontaneous play 0.00 0.02 0.06 0.21 0.16 0.30 0.25 04 Factor analysis
Get eigenvalues for determining number of factors.
att.mat <- select(wide.attitudes, -workerid)
ev <- eigen(cor(x=att.mat)) # get eigenvalues
ap <- parallel(subject=nrow(att.mat), var=ncol(att.mat),
rep=100,cent=.05)
nS <- nScree(x=ev$values, aparallel=ap$eigen$qevpea)
plotnScree(nS)
Now plot factor analysis
Because we predicted 3 factors, I am going with 3 here (confirmatory FA)
In the graph items are organized by category, from Early/Active Learning -> Affection and Attachment -> Rules and Respect.
Things are looking a bit hazy in terms of individual items mapping onto factors. In particular, most of the Early/Active Learning AND Affection and Attachment items map onto Factor 1. Factor 2 has an almost even mix of all 3 categories. Also, it seems like items related to Rules (Factor 3) are separate from those ostensibly related to Respect (Factor 2). It seems that items related to strict rules are most consistently hanging together across experiments.
n.factors <- 3
af <- factanal(x=att.mat, factors = n.factors, scores = "regression")
loadings <- data.frame(af$loadings[,1:n.factors]) %>%
mutate(item = rownames(af$loadings)) %>%
gather(factor, loading, starts_with("Factor"))
loadings$item <- factor(loadings$item, levels = c("respect parents teachers","punish misbehavior","dont interrupt","wait when told","control behavior", "children should be grateful", "strict rules", "punish for small rules", "worry about misbehavior", "dont impose structure","rules prevent tantrums", "kids should not decide", "strong bond mom", "safe loving environment", "holding and cradling", "talk about opinions", "give children comfort", "talk about emotions", "talk about feelings before mature", "not spoiled with too much attention", "too much affection does not harm", "control emotions if comforted", "examples are good", "can learn good and bad ", "follow-in good", "spontaneous play", "learn before talking", "learn by playing", "learn contingencies", "learn from repetetive behaviors", "play pretend", "simple toys", "reasons for rules", "learn math before school", "learn from other children", "play with friends", "new experiences", "reading before school", "nonsense games", "play freely", "parents choose toys", "reading before speaking"))
labels$item <- labels$short_sent
loadings <- join(loadings, labels, by = "item")
qplot(factor, item, fill=loading, geom="tile", data = loadings) +
scale_fill_continuous(low="#000000", high="#FFFFFF") + facet_grid(. ~ category)
qplot(factor, item, fill=loading > .35, geom="tile", data = loadings) +
scale_fill_solarized()+ facet_grid(. ~ category)
Now with the suggested 5 factors.
n.factors <- 5
af <- factanal(x=att.mat, factors = n.factors, scores = "regression")
loadings <- data.frame(af$loadings[,1:n.factors]) %>%
mutate(item = rownames(af$loadings)) %>%
gather(factor, loading, starts_with("Factor"))
loadings$item <- factor(loadings$item, levels = c("respect parents teachers","punish misbehavior","dont interrupt","wait when told","control behavior", "children should be grateful", "strict rules", "punish for small rules", "worry about misbehavior", "dont impose structure","rules prevent tantrums", "kids should not decide", "strong bond mom", "safe loving environment", "holding and cradling", "talk about opinions", "give children comfort", "talk about emotions", "talk about feelings before mature", "not spoiled with too much attention", "too much affection does not harm", "control emotions if comforted", "examples are good", "can learn good and bad ", "follow-in good", "spontaneous play", "learn before talking", "learn by playing", "learn contingencies", "learn from repetetive behaviors", "play pretend", "simple toys", "reasons for rules", "learn math before school", "learn from other children", "play with friends", "new experiences", "reading before school", "nonsense games", "play freely", "parents choose toys", "reading before speaking"))
loadings <- join(loadings, labels, by = "item")
qplot(factor, item, fill=loading, geom="tile", data = loadings) +
scale_fill_continuous(low="#000000", high="#FFFFFF")+ facet_grid(. ~ category)
qplot(factor, item, fill=loading > .35, geom="tile", data = loadings) +
scale_fill_solarized()+ facet_grid(. ~ category)+ facet_grid(. ~ category)
Merge back in subject id info.
a.scores <- af$scores %>%
data.frame %>%
mutate(workerid = as.character(wide.attitudes$workerid)) %>%
left_join(subinfo)And summary:
4.1 Factors by SES
a.factor.names <- c("Active Learning and Affection", "Early Learning, Affection, Respect", "Strict Rules")
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(ses != "") %>%
mutate(high.ses = ses > 5,
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(high.ses, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=high.ses,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
4.2 Factors by Gender
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(gender != "") %>%
mutate(male.gender = gender == "Male",
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(male.gender, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=male.gender,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
4.3 Factors by Education
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(education != "") %>%
mutate(high.education = (education == "4year" |education == "someGrad" | education == "Grad"),
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(high.education, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=high.education,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
4.4 Factors by Age
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(age != "") %>%
mutate(young.age = (age == "0-19" |age == "20-29"),
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(young.age, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=young.age,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
4.5 Factors by Children
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(children != "") %>%
mutate(n.children = children,
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(n.children, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=n.children,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(children != "") %>%
mutate(yes.children = (children == "1" | children == "2" | children == "3"| children == "4"| children == "5"),
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(yes.children, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=yes.children,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
4.6 Factors by Ethnicity
ms <- a.scores %>%
gather(factor, score, starts_with("Factor")) %>%
filter(ethnicity != "") %>%
mutate(hispanic = ethnicity== "Hispanic",
factor.num = as.numeric(str_replace(factor,"Factor","")),
factor.name = a.factor.names[factor.num]) %>%
group_by(hispanic, factor.name) %>%
multi_boot_standard(col = "score")
qplot(factor.name, mean, fill=hispanic,
geom = "bar", stat = "identity", position = "dodge",
data=ms) +
geom_linerange(aes(ymin = ci_lower, ymax = ci_upper),
position = position_dodge(width = .9))
5 Conclusions
Our predicted factor structure did not hold up well. The items that consistently hang together and form their own factor are those related to strict rules.
Might need to rethink the organization of attitudes. Something separating out parents who see their kids as “projects” versus believing they will naturally thrive might make sense. The grouping of early learning and affection items into one factor might reflect this latent theory.