Introduction
Approximately one in three Americans have a criminal record (ACLU, 2017). It is known that stable employment can help reduce recidivism, yet 75% of justice-involved individuals (JIIs) remain unemployed a year post-release (Nally et al., 2014; Visher et al., 2011). Some claim that JIIs are loyal and hardworking when given an opportunity to make a living but there is little empirical evidence to support it (ACLU, 2017). Our team conducted a research study to test whether organizational commitment, or an employee’s feeling of dedication and responsibility toward an organization (Klein et al., 2012), and their intentions to quit their job differs between JIIs and non-JIIs. This study consisted of 167 survey responses from 91 JIIs and 76 non-JIIs collected through community-based organizations and Prolific (an online survey data collection platform). We measured organizational commitment in two ways: Klein et al (2012) conceptualization of commitment (KUT) and the Three-Component Model of Commitment (TCM; Meyer & Allen, 1991). Turnover intentions were measured using a scale adapted based on Hom et al. (1984).
Our hypotheses are informed by Social Exchange Theory, Social Identity Theory, and what previous literature is available related to this topic (Avanzi et al., 2014, Ellmers et al., 2002; Griffith et al. 2019; Tajfel & Turner, 1979). For exploratory purposes, we also measured Global Job Embeddedness (JE; Crossley et al., 2007), Perceived Stigma at Work (PSS; James et al, 1994) and included several open-ended questions intended to build upon the work of Anazodo et al. (2019). This study fills a gap in the literature as the only study that we are aware of to explore job attitudes of justice-involved individuals.
Sample Demographics
This sample was 62% male and 37% females. The ages of the sample range from 20 to 65 with a median age of 39. The majority of respondents were white (79%) followed by Black (12%), Hispanic/Latino/Spanish (5%), and Asian (3%) participants. Education in this sample was varied with 28% having a high school degree or equivalent, 24% had some college, 22% had a bachelors degree, 18% had a certificate or associate’s degree, and 5% had a graduate degree. In terms of relationships, 53% were married or living with a partner, 31% were never married, and 10% were divorced.
JIIs made up 54% of the sample. Of the JII sample, 57% had a felony
charge while 42% had a misdemeanor. 76% of the JIIs had been
incarcerated for some amount of time.
In terms of transparency at work 81% of the sample said their employer
was aware of their background, 72% said their supervisor was aware of
their background, and 56% said their peers were aware of their
background.
JII_cat <- c("JII", "severity", "incar", "emp_aware", "sup_aware", "peer_aware")
demos <- c("gender", "race", "edu", "mar_stat")
# Frequency tables grouped by JII
for (var in demos) {
# Generate frequency table grouped by JII using summarytools
freq_table <- freq(data[[var]], by = data$JII, order = "freq")
# Extract the frequency table and convert to data frame
freq_df <- as.data.frame(freq_table)
# Print the table with kable
print(kable(freq_df, caption = paste("Frequency Table for", var, "Grouped by JII")))
cat("\n\n")
}##
##
## Table: Frequency Table for gender Grouped by JII
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:-----------------------|----:|-----------:|------------:|-----------:|------------:|
## |Male | 104| 62.2754491| 62.27545| 60.8187135| 60.81871|
## |Female | 62| 37.1257485| 99.40120| 36.2573099| 97.07602|
## |Prefer to self-descirbe | 1| 0.5988024| 100.00000| 0.5847953| 97.66082|
## |<NA> | 4| NA| NA| 2.3391813| 100.00000|
## |Total | 171| 100.0000000| 100.00000| 100.0000000| 100.00000|
##
##
##
##
## Table: Frequency Table for race Grouped by JII
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:--------------------------------|----:|-----------:|------------:|-----------:|------------:|
## |White | 124| 78.9808917| 78.98089| 72.5146199| 72.51462|
## |Black or African American | 19| 12.1019108| 91.08280| 11.1111111| 83.62573|
## |Hispanic, Latino, or Spanish | 8| 5.0955414| 96.17834| 4.6783626| 88.30409|
## |Asian | 4| 2.5477707| 98.72611| 2.3391813| 90.64327|
## |American Indian or Alaska Native | 1| 0.6369427| 99.36306| 0.5847953| 91.22807|
## |Prefer to self-descirbe | 1| 0.6369427| 100.00000| 0.5847953| 91.81287|
## |<NA> | 14| NA| NA| 8.1871345| 100.00000|
## |Total | 171| 100.0000000| 100.00000| 100.0000000| 100.00000|
##
##
##
##
## Table: Frequency Table for edu Grouped by JII
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:--------------------------------------------------------------------------------------------|----:|----------:|------------:|----------:|------------:|
## |High school degree or equivalent (GED) | 47| 28.143713| 28.14371| 27.485380| 27.48538|
## |Some college | 40| 23.952096| 52.09581| 23.391813| 50.87719|
## |Bachelor's degree | 37| 22.155689| 74.25150| 21.637427| 72.51462|
## |Certificate or Associate degree including trade school or professional certification program | 30| 17.964072| 92.21557| 17.543860| 90.05848|
## |Graduate degree | 9| 5.389222| 97.60479| 5.263158| 95.32164|
## |Less than a high school diploma | 4| 2.395210| 100.00000| 2.339181| 97.66082|
## |<NA> | 4| NA| NA| 2.339181| 100.00000|
## |Total | 171| 100.000000| 100.00000| 100.000000| 100.00000|
##
##
##
##
## Table: Frequency Table for mar_stat Grouped by JII
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:------------------------------|----:|----------:|------------:|----------:|------------:|
## |Married or living with partner | 88| 53.012048| 53.01205| 51.461988| 51.46199|
## |Never married | 52| 31.325301| 84.33735| 30.409357| 81.87135|
## |Divorced | 17| 10.240964| 94.57831| 9.941520| 91.81287|
## |Separated | 7| 4.216868| 98.79518| 4.093567| 95.90643|
## |Widowed | 2| 1.204819| 100.00000| 1.169591| 97.07602|
## |<NA> | 5| NA| NA| 2.923977| 100.00000|
## |Total | 171| 100.000000| 100.00000| 100.000000| 100.00000|# Frequency tables without grouping
for (var in JII_cat) {
# Generate frequency table without grouping using summarytools
freq_table <- freq(data[[var]], order = "freq")
# Extract the frequency table and convert to data frame
freq_df <- as.data.frame(freq_table)
# Print the table with kable
print(kable(freq_df, caption = paste("Frequency Table for", var)))
cat("\n\n")
}##
##
## Table: Frequency Table for JII
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:----------------------|----:|---------:|------------:|----------:|------------:|
## |Justice Involved | 91| 54.49102| 54.49102| 53.216374| 53.21637|
## |No Criminal Background | 76| 45.50898| 100.00000| 44.444444| 97.66082|
## |<NA> | 4| NA| NA| 2.339181| 100.00000|
## |Total | 171| 100.00000| 100.00000| 100.000000| 100.00000|
##
##
##
##
## Table: Frequency Table for severity
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:-----------|----:|---------:|------------:|---------:|------------:|
## |Felony | 52| 57.14286| 57.14286| 30.40936| 30.40936|
## |Misdemeanor | 39| 42.85714| 100.00000| 22.80702| 53.21637|
## |<NA> | 80| NA| NA| 46.78363| 100.00000|
## |Total | 171| 100.00000| 100.00000| 100.00000| 100.00000|
##
##
##
##
## Table: Frequency Table for incar
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:-----|----:|---------:|------------:|---------:|------------:|
## |Yes | 70| 76.92308| 76.92308| 40.93567| 40.93567|
## |No | 21| 23.07692| 100.00000| 12.28070| 53.21637|
## |<NA> | 80| NA| NA| 46.78363| 100.00000|
## |Total | 171| 100.00000| 100.00000| 100.00000| 100.00000|
##
##
##
##
## Table: Frequency Table for emp_aware
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:------|----:|----------:|------------:|----------:|------------:|
## |Yes | 74| 81.318681| 81.31868| 43.274854| 43.27485|
## |Unsure | 9| 9.890110| 91.20879| 5.263158| 48.53801|
## |No | 8| 8.791209| 100.00000| 4.678363| 53.21637|
## |<NA> | 80| NA| NA| 46.783626| 100.00000|
## |Total | 171| 100.000000| 100.00000| 100.000000| 100.00000|
##
##
##
##
## Table: Frequency Table for sup_aware
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:------|----:|---------:|------------:|----------:|------------:|
## |Yes | 66| 72.52747| 72.52747| 38.596491| 38.59649|
## |Unsure | 13| 14.28571| 86.81319| 7.602339| 46.19883|
## |No | 12| 13.18681| 100.00000| 7.017544| 53.21637|
## |<NA> | 80| NA| NA| 46.783626| 100.00000|
## |Total | 171| 100.00000| 100.00000| 100.000000| 100.00000|
##
##
##
##
## Table: Frequency Table for peer_aware
##
## | | Freq| % Valid| % Valid Cum.| % Total| % Total Cum.|
## |:------|----:|---------:|------------:|----------:|------------:|
## |Yes | 51| 56.04396| 56.04396| 29.824561| 29.82456|
## |No | 24| 26.37363| 82.41758| 14.035088| 43.85965|
## |Unsure | 16| 17.58242| 100.00000| 9.356725| 53.21637|
## |<NA> | 80| NA| NA| 46.783626| 100.00000|
## |Total | 171| 100.00000| 100.00000| 100.000000| 100.00000|Descriptive Statistics
The average of the unidimensional commitment score (KUT) had an average of 3.65 (a = .95), the average of affective commitment (ACS) was 4.48, continuance commitment (CCS) was 4.6, and normative commitment (NCS) was 4.23. The overall alpha for the three forms of commitment (TCM) was a = .84. The average intention to quit (INQ) was 3.47 (a = .90). The skewness and kurtosis for all of these variables was between +/-2.
NCS and KUT are significantly negatively correlated with INQ (-.59; -.62).
Intentions to quit were significantly negatively correlated with all measures of commitment (KUT, TCM, ACS, and NCS) except for continuance commitment.
Exploratory variable of perceived stigma was positively correlated with age (.16) and CCS (.16) so that more stigma is associated with greater age and more continuance commitment.
Exploratory variable of job embeddedness is significantly positively correlated with all measures of commitment and significantly negatively correlated with intentions to quit.
The other variables in the descriptive statistics table are not detailed here because they were exploratory variables and analysis for these variables have yet to be conducted. Further, our research team is in the process collecting additional data so that more complex analyses can be done.
Desc_table <- data %>%
select(age, KUT_mean, TCM_mean, ACS_mean, CCS_mean, NCS_mean, INQ_mean, PSS_mean, JE_mean)
# Compute descriptive statistics by group
describe_stats <- describe(Desc_table)
# Convert describe_stats to a data frame for Word table
describe_stats_df <- as.data.frame(describe_stats)
# Display descriptive statistics in a kable table
describe_stats_df %>%
kable("html", caption = "Descriptive Statistics") | vars | n | mean | sd | median | trimmed | mad | min | max | range | skew | kurtosis | se | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| age | 1 | 167 | 40.245509 | 10.251340 | 39.000000 | 39.962963 | 10.37820 | 20 | 65.0 | 45.0 | 0.2616282 | -0.6156513 | 0.7932725 |
| KUT_mean | 2 | 169 | 3.649408 | 1.049135 | 3.750000 | 3.728102 | 1.11195 | 1 | 5.0 | 4.0 | -0.4459277 | -0.5480210 | 0.0807027 |
| TCM_mean | 3 | 169 | 4.425296 | 1.356362 | 4.625000 | 4.466241 | 1.48260 | 1 | 7.0 | 6.0 | -0.3015361 | -0.4927695 | 0.1043355 |
| ACS_mean | 4 | 169 | 4.476331 | 1.886521 | 5.000000 | 4.583942 | 2.22390 | 1 | 7.0 | 6.0 | -0.4155754 | -1.0158031 | 0.1451170 |
| CCS_mean | 5 | 169 | 4.589744 | 1.652319 | 5.000000 | 4.666667 | 1.48260 | 1 | 7.0 | 6.0 | -0.4010410 | -0.9348435 | 0.1271015 |
| NCS_mean | 6 | 169 | 4.226825 | 1.889953 | 4.333333 | 4.282238 | 2.47100 | 1 | 7.0 | 6.0 | -0.2283308 | -1.1184022 | 0.1453810 |
| INQ_mean | 7 | 169 | 3.467456 | 1.974815 | 3.333333 | 3.357664 | 2.47100 | 1 | 7.0 | 6.0 | 0.2590373 | -1.2107619 | 0.1519088 |
| PSS_mean | 8 | 167 | 3.225150 | 1.351337 | 3.200000 | 3.158519 | 1.18608 | 1 | 7.6 | 6.6 | 0.5480121 | 0.0049728 | 0.1045696 |
| JE_mean | 9 | 169 | 3.206255 | 1.083925 | 3.285714 | 3.245047 | 1.05900 | 1 | 5.0 | 4.0 | -0.3082918 | -0.6981033 | 0.0833788 |
# Reliability coefficients
# Compute alpha coefficients
alpha_results <- data.frame(
Scale = c("KUT", "TCM", "INQ", "PSS", "JE"),
Alpha = c(
psych::alpha(data[, c("KUT_1", "KUT_2", "KUT_3", "KUT_4")])$total$raw_alpha,
psych::alpha(data[, c("TCM_1", "TCM_2", "TCM_3", "TCM_4", "TCM_5", "TCM_6", "TCM_7", "TCM_8")])$total$raw_alpha,
psych::alpha(data[, c("INQ_1", "INQ_2", "INQ_3")])$total$raw_alpha,
psych::alpha(data[, c("PSS_1", "PSS_2", "PSS_3", "PSS_4", "PSS_5")])$total$raw_alpha,
psych::alpha(data[, c("JE_1", "JE_2", "JE_3", "JE_4", "JE_5", "JE_6r", "JE_7")])$total$raw_alpha
)
)
# Create a table
alpha_results %>%
kable("html", caption = "Cronbach's Alpha Coefficients")| Scale | Alpha |
|---|---|
| KUT | 0.9541303 |
| TCM | 0.8355125 |
| INQ | 0.9032916 |
| PSS | 0.7715280 |
| JE | 0.9167611 |
# Compute correlation table for key variables
cor_results <- apa.cor.table(Desc_table)
print(cor_results)##
##
## Means, standard deviations, and correlations with confidence intervals
##
##
## Variable M SD 1 2 3 4
## 1. age 40.25 10.25
##
## 2. KUT_mean 3.65 1.05 .14
## [-.01, .29]
##
## 3. TCM_mean 4.43 1.36 .13 .77**
## [-.02, .27] [.70, .83]
##
## 4. ACS_mean 4.48 1.89 .21** .81** .77**
## [.07, .36] [.76, .86] [.70, .83]
##
## 5. CCS_mean 4.59 1.65 .03 .18* .60** .10
## [-.12, .18] [.03, .32] [.50, .69] [-.05, .25]
##
## 6. NCS_mean 4.23 1.89 .08 .78** .87** .73**
## [-.07, .23] [.72, .84] [.83, .90] [.65, .79]
##
## 7. INQ_mean 3.47 1.97 -.14 -.62** -.58** -.70**
## [-.28, .02] [-.71, -.52] [-.67, -.47] [-.77, -.61]
##
## 8. PSS_mean 3.23 1.35 .16* -.02 .06 -.00
## [.01, .31] [-.17, .14] [-.10, .21] [-.16, .15]
##
## 9. JE_mean 3.21 1.08 .11 .75** .87** .75**
## [-.04, .26] [.68, .81] [.82, .90] [.68, .81]
##
## 5 6 7 8
##
##
##
##
##
##
##
##
##
##
##
##
##
##
## .21**
## [.06, .35]
##
## -.07 -.59**
## [-.22, .08] [-.68, -.48]
##
## .16* -.03 -.02
## [.01, .31] [-.19, .12] [-.17, .13]
##
## .41** .80** -.60** .03
## [.28, .53] [.74, .85] [-.69, -.49] [-.12, .18]
##
##
## Note. M and SD are used to represent mean and standard deviation, respectively.
## Values in square brackets indicate the 95% confidence interval.
## The confidence interval is a plausible range of population correlations
## that could have caused the sample correlation (Cumming, 2014).
## * indicates p < .05. ** indicates p < .01.
## Relevant Visualizations
Various visualizations were created to get a sense of the distributions of various variables of interest. While the box plots don’t show any concerning outliers, there are some signs of skewness and non-normality in various histograms and Q-Q plots.
Box Plots
# Generating Multiple Box plots
# Loop through numeric variables and create boxplots
for (var in names(Desc_table)) {
# Create the boxplot
p <- ggplot(data, aes(y = .data[[var]])) +
geom_boxplot(fill = "lightblue", color = "black") +
labs(title = paste("Boxplot of", var), y = var) +
theme_minimal()
print(p)
}
Histograms
hyp_variables <- c("TCM_mean", "KUT_mean", "INQ_mean", "ACS_mean", "NCS_mean", "CCS_mean")
# Create histograms in a loop
for (var in hyp_variables) {
print(
ggplot(data, aes_string(x = var)) +
geom_histogram(aes(y = ..density..), binwidth = 1, fill = "skyblue", color = "black") +
geom_density(color = "red", size = 1) +
labs(title = paste("Scale Score Distribution of", var), x = "Value", y = "Density") +
theme_minimal()
)
}
Hypothesis Testing
Our hypotheses are as follows:
H1. We expect JIIs will have lower turnover intentions compared to other employees.
H2. We expect JIIs to have significantly greater organizational commitment.
H3. Criminal background will moderate the relationship between organizational commitment and turnover intentions, such that those with a criminal background will have a stronger relationship between organizational commitment and turnover intentions.
T-Tests
H1: There was no significant difference in turnover intentions between JIIs and other employees (p = 0.77).
H2: JIIs (M = 3.82) had significantly higher organizational commitment when looking at the general measure of commitment (KUT) than other employees (M = 3.45). This was a small to medium effect (d = -0.36).
JIIs (M = 4.6) also had significantly higher normative commitment (p = 0.005) than other employees (M = 3.79). This was a small to medium effect (d = -0.44 ).
# Perform Shapiro-Wilk normality test for each variable
shapiro_results <- lapply(hyp_variables, function(var) {
test <- shapiro.test(data[[var]])
data.frame(Variable = var, W = unname(test$statistic), p_value = test$p.value)
})
# Combine results into a single data frame
shapiro_df <- bind_rows(shapiro_results)
# Print results in a kable table
shapiro_df %>%
kable("html", caption = "Shapiro-Wilk Normality Test Results")| Variable | W | p_value |
|---|---|---|
| TCM_mean | 0.9829905 | 0.0366288 |
| KUT_mean | 0.9397125 | 0.0000015 |
| INQ_mean | 0.9156747 | 0.0000000 |
| ACS_mean | 0.9242418 | 0.0000001 |
| NCS_mean | 0.9400743 | 0.0000016 |
| CCS_mean | 0.9449078 | 0.0000039 |
# T-tests loop
t_test_results <- lapply(hyp_variables, function(var) {
# Perform t-test
t_test <- t.test(data[[var]] ~ data$JII)
# Return a data frame with results
t_statistic <- t_test$statistic
p_value <- t_test$p.value
conf_int <- paste(round(t_test$conf.int[1], 3), "to", round(t_test$conf.int[2], 3)) # Format confidence interval
df <- t_test$parameter # Degrees of freedom
JII_means <- paste(round(t_test$estimate[1], 3))
Non_JII_means<- paste(round(t_test$estimate[2], 3))
stderr <- round(t_test$stderr, 3) # Standard error of the difference
# Return a data frame with results
data.frame(
Variable = var,
t_statistic = t_statistic,
p_value = p_value,
Confidence_Interval = conf_int,
Degrees_of_Freedom = df,
JII_Means = JII_means,
Non_JII_Means = Non_JII_means,
Standard_Error = stderr
)
})
# Combine results into a single data frame
t_test_df <- bind_rows(t_test_results)
rownames(t_test_df) <- NULL
# Print results in a kable table
t_test_df %>%
kable("html", caption = "T-Test Results")| Variable | t_statistic | p_value | Confidence_Interval | Degrees_of_Freedom | JII_Means | Non_JII_Means | Standard_Error |
|---|---|---|---|---|---|---|---|
| TCM_mean | 1.2834976 | 0.2011257 | -0.144 to 0.679 | 164.2416 | 4.545 | 4.278 | 0.208 |
| KUT_mean | 2.3210724 | 0.0215272 | 0.055 to 0.68 | 162.2460 | 3.821 | 3.454 | 0.158 |
| INQ_mean | 0.2810775 | 0.7790259 | -0.522 to 0.695 | 154.9661 | 3.495 | 3.408 | 0.308 |
| ACS_mean | 1.7293521 | 0.0856748 | -0.071 to 1.074 | 159.9942 | 4.725 | 4.224 | 0.290 |
| NCS_mean | 2.8066317 | 0.0056359 | 0.238 to 1.368 | 158.1872 | 4.597 | 3.794 | 0.286 |
| CCS_mean | -1.6705424 | 0.0967555 | -0.927 to 0.077 | 160.9550 | 4.374 | 4.798 | 0.254 |
# Loop through each variable in hyp_variables
for (var in hyp_variables) {
# Exclude rows where the current variable or JII is NA
group_means <- data %>%
filter(!is.na(.data[[var]]), !is.na(JII)) %>% # Remove NA values for both variable and JII
group_by(JII) %>%
summarise(Mean = mean(.data[[var]], na.rm = TRUE), .groups = "drop")
# Create the bar chart showing means grouped by JII
plot <- ggplot(group_means, aes(x = JII, y = Mean, fill = JII)) +
geom_bar(stat = "identity", color = "black", width = 0.6, alpha = 0.7) +
geom_text(aes(label = round(Mean, 2)), vjust = -0.3) + # Add mean labels on top of the bars
labs(title = paste("Mean of", var, "by JII Group"),
y = "Mean Value", x = "JII Group") +
theme_minimal() +
scale_fill_manual(values = c("blue", "gold")) # Customize colors
# Print the plot
print(plot)
}
# Cohen's d
data$JII <- as.factor(data$JII)
# Compute Cohen's d
cohen_d_KUT <- cohen.d(data$KUT_mean, data$JII)
cohen_d_NCS <- cohen.d(data$NCS_mean, data$JII)
# Extract the middle value from cohen.d
cohen_d_KUT_value <- cohen_d_KUT$cohen.d[2] # Middle value
cohen_d_NCS_value <- cohen_d_NCS$cohen.d[2] # Middle value
# Create a table with the extracted values
effect_table <- data.frame(
Variable = c("KUT_mean", "NCS_mean"),
Cohen_d = c(cohen_d_KUT_value, cohen_d_NCS_value)
)
# Display table in HTML format
kable(effect_table, format = "html", caption = "Cohen's d Effect Size Coefficients")| Variable | Cohen_d |
|---|---|
| KUT_mean | -0.3611835 |
| NCS_mean | -0.4398060 |
Moderated Linear Regression
Various plots were generated to check the assumptions of linear regression which can be viewed below. A moderated linear regression model was created to predict INQ with KUT mean score, JII status, and an interaction term as the predictors. While the model was significant (R2 = .406), only KUT_mean (b = -1.45) was a significant predictor of INQ explaining 24% of the variance. The results of the analysis suggest that JIIs do not have a significantly different INQ than non-JIIs and that JII status does not moderate the relationship between KUT and INQ.
data <- data %>%
mutate(JII_dummy = ifelse(JII == "Justice Involved", 1, 0))
JII_model <- lm(INQ_mean ~ KUT_mean + JII_dummy + KUT_mean*JII_dummy, data = data)
check_model(JII_model)
##
##
## Regression results using INQ_mean as the criterion
##
##
## Predictor b b_95%_CI sr2 sr2_95%_CI Fit
## (Intercept) 8.41** [7.15, 9.67]
## KUT_mean -1.45** [-1.80, -1.10] .24 [.14, .35]
## JII_dummy -0.99 [-2.73, 0.75] .00 [-.01, .02]
## KUT_mean:JII_dummy 0.42 [-0.04, 0.89] .01 [-.01, .04]
## R2 = .406**
## 95% CI[.29,.49]
##
##
## Note. A significant b-weight indicates the semi-partial correlation is also significant.
## b represents unstandardized regression weights.
## sr2 represents the semi-partial correlation squared.
## Square brackets are used to enclose the lower and upper limits of a confidence interval.
## * indicates p < .05. ** indicates p < .01.
## Conclusion
Organizations care about employee commitment because of its impact on outcomes such as satisfaction, job involvement, and turnover (Meyer et al., 2002). We found that JIIs have significantly higher organizational commitment (KUT) and normative commitment compared to non-JIIs. While we did not find a significant difference in JIIs intentions to quit it is worth noting that intentions to quit were significantly negatively correlated with Normative commitment (-.59) as well as the KUT measure of commitment (-.62). These findings suggest that JIIs may feel that their commitment is the right thing to do when an organization provides an opportunity that few others would (Avanzi et al., 2014). This directly aligns with and supports the claim discussed above about the potential loyalty of justice-involved workers. This research fills a gap in literature related to the potentially unique job attitudes of JIIs and may help inform hiring decisions that could benefit both organizations and a population in need.