Data from: Striving and thriving: Gender differences in the effects of climbing the socioeconomic ladder on stress and discrimination
Data files
Jun 11, 2026 version files 1.42 MB
-
boots.ado
1.59 KB
-
boots.hlp
1.05 KB
-
Figures_1_and_2.do
3.90 KB
-
Infile_riseup-stress.do
322 B
-
README.md
17.02 KB
-
riseup-stress-public.csv
836.41 KB
-
riseup-stress-public.dta
541.56 KB
-
Table_1.do
536 B
-
Table_2.do
196 B
-
Table_3_ASQ.do
6.07 KB
-
Table_3_PSS.do
4.98 KB
-
Table_4.do
1.77 KB
-
Table_5_sensitivity.do
3.98 KB
-
Table_5.do
3.98 KB
Abstract
Background: Recent evidence suggests upward socioeconomic mobility (“striving”) may result in worse physical health. It has been hypothesized that striving might result in greater stress and discrimination, but prior observational evidence of this has been mixed.
Methods: We analyzed data from The RISE UP study, a natural experimental study of 1270 low-income students in Los Angeles who were offered entrance to high-performing charter high schools via a random lottery. We compared lottery winners to those waitlisted, enabling objective assessment of striving with less confounding and selection bias. Self-reported stress and discrimination were collected prospectively from adolescence into early adulthood.
Results: Striving was associated with increased self-reported psychological stress, particularly among females, but not with greater perceived discrimination. Gender differences in stress emerged most strongly during early adulthood, with females exposed to high-performing schools reporting significantly higher stress, while males reported lower stress. No significant associations were observed between striving and perceived discrimination by lottery status for either gender, though females consistently reported greater overall discrimination than males.
Conclusions: These findings support the weathering hypothesis, suggesting that the process of upward mobility may incur physiological and psychological costs, especially for females. The study highlights the need for further inquiry into mechanisms underlying gender disparities in the health effects of striving, and underscores the complex interplay between socioeconomic mobility, stress, discrimination, and health.
Dataset DOI: 10.5061/dryad.95x69p8z8
Descriptions
Datafiles
Data files for all analyses: riseup-stress-public. File format in .csv or Stata (.dta) provided.
File: riseup-stress-public.dta
Description: RISE Up dataset for stress study in Stata format
File: riseup-stress-public.csv
Description: RISE Up dataset for stress study in csv format
Data structure
The data structure is in long format.
- id_stress: anonymized subject id
- SCH09ID_stress: anonymized unique id# for the high school each subject attended in 9th grade
- time: Time references the survey wave of the study. 0=baseline; 1=10th grade; 2=11th grade; 3=12th grade; 4=age 20; 5=age21; 6=age 22
Description of the study, data, and file structure
The Reducing Inequities through Social and Educational Change (RISE Up) Study is a natural experiment comparing a cohort of mostly Latinx adolescents from low-income neighborhoods of Los Angeles who applied to high-performing, public, charter high schools in 2013 and 2014. The study examines those who applied to these high performing schools, comparing lottery winners vs. lottery waitlsted students. Prior results of this study have found that lottery winners had lower rates of substance use and delinquent behaviors compared to those waitlisted. We also found that lottery winners had lower rates of self-reported obesity and poor physical health, but only among males. Among females, lottery winners had much higher rates of obesity and poor physical health. Thus, we hypothesized that exposure to high performing schools would lead to more stress among females, but lead to lower or no differences in stress among males. We hypothesized similar patterns regarding self-reported discrimination.
Our results regarding stress were consistent with our hypothesis. We found that among females, lottery winners reported more stress than lottery waitlisters. Among males, lottery winners reported less stress than lottery waitlisters. We found no differences in self-reported discrimination among lottery winners vs. waitlisted among males or females.
We recruited 1270 adolescents who applied to 5 high-performing charter schools in Los Angeles in 2013 and 2014 (two consecutive academic years). Eligible subjects were those admitted to these schools through their random admission lottery. We recruited a random subset of lottery winners and those waitlisted with the goal of obtaining equal numbers of winners and waitlisted students from each of the 5 schools. We studied high-performing charter high schools that were over-subscribed (more applicants than seats available) and admitted students based on a random lottery. This allowed us to recruit applicants to these schools for participation in our study and observe comparable groups of low-income students exposed to higher- and lower-performing schools. Subjects were recruited at entry into 9th grade and surveyed every 1-2 years through age 22. The study is ongoing with plans to continue following subjects through 2026 (age 26).
This dataset includes baseline student and parental demographics, self-reported stress and discrimination, and baseline grade point average derived from middle school transcripts. We compared lottery winners to those waitlisted (EXP1=1 vs 0, respectively). Self-reported stress and discrimination were collected prospectively from adolescence into early adulthood at each time point of the survey (time=0-6). time 0=baseline, time1=grade 10, time2=grade11, time3=grade12, time4=age20, time5=age21 time6=age22. We used hierachical models to examine repeated measures of stress over time for lottery winners/waitlisted students stratified by gender. The hierarchical models adjust for multiple measures of the outcomes over time and students clustered by school. We examined stress during adolescence (time=0-4) using the adolescent stress questionnaire (ASQ) and stress during early adulthood (time=5-6) using the perceived stress scale. Stress was not measured at time=3 (grade 12). We also examined perceived discrimination at time=5 using the everyday discrimination scale. Our analyses adjusted for sociodemographic measures were gathered in the baseline survey and were all self-reported by the adolescent. These included gender (male, female, other), race/ethnicity, country of birth, native language, and parental demographics (country of birth, employment, and education attainment). We use the term gender rather than sex because that is how we asked the question in our survey. Given the distribution by race/ethnicity in the sample, respondents were categorized as Latinx versus non-Latinx. We adjusted for middle school grade point average derived from school transcripts and self-report. We also adjusted for the combination of charter schools that students applied to, referred to as the risk set (criskset).
List of variables
- id_stress: anonymized unique id number for each study participant
- SCH09ID_stress: anonymized unique id# for the high school each subject attended in 9th grade
- time: Time references the survey wave of the study. 0=baseline; 1=10th grade; 2=11th grade; 3=12th grade; 4=age 20; 5=age21; 6=age 22
- male: self-reported male sex- female=0 male=1
- usborn: US born- 0=no; 1=yes
- engnat: English is native language- 0=no; 1=yes
- r_latino: Latino ethnicity- 0=no; 1=yes
- p_usborn: At least 1 parent born in US- 0=no; 1=yes
- p_hsgrad: At least 1 parent hs graduate- 0=no; 1=yes
- p_workfull: At least 1 parent working full time- 0=no; 1=yes
- EXP1: SAMPLE original sampling frame 0=lottery waitlist; 1=lottery winner. Subjects were recruited from applicant list of 5 high performing charter schools in Los Angeles that were oversubscribed (more applicants than seats available). The admission lottery was random. We sampled students from the top of the list who were offered seats (lottery winners) and from the bottom of the list (lottery waitlisted students).
- criskset: Risk set of schools applied to plus study cohort. Categorical indicator for different combination of the 5 charter schools involved in the original sampling. The original cohort were sample over 2 academic years 2013-14 and 2014-15. The first column represents cohort: 1=2013-2014; 2=2014-15. Columns 2-6 represent each charter school (column 2=charter#1; column 3=charter#2, etc...) of the sampling frame and whether the subject applied for admission to that that charter school 0=no; 1=yes. Thus, "100001" represents a subject who was recruited in 2013-2014 and applied to charter#5 only. "200110" represents a subject who was recruited in 2014-15 and applied to charter school#3 and charter school#4. A few combination of charter school applications had small cell sizes and were collapsed into one group (99999, n=27). The reason to include criskset as a covariate in the analyses is that subjects who applied to more than 1 charter school had more opportunities to be admitted to 1 of the 5 charter schools. Also each charter school had different # of applicants. Thus, winning the admission lottery was easier at some schools than others.
- gpa8comb: Middle school transcript gpa with self-report if missing, categorical. 2=GPA<2.0; 3=2.0-2.5; 4=2.6-3.0; 5=3.1-3.5; 6=3.6-4.0; 9=missing transcript and self-reported GPA.
Discrimination variables
The measure of discrimination was derived from the Everyday Discrimination Scale. Details can be accessed here: https://hsph.harvard.edu/research/williams-group/measuring-discrimination-resource/
Citation for the full scale:
Williams DR, Yu Y, Jackson JS, Anderson NB. Racial Differences in Physical and Mental Health. J Health Psychol. 1997;2(3):335-351. doi:10.1177/135910539700200305
We asked an abbreviated version of the original scale:
In your day-to-day life, how often do any of the following things happen to you?
- You are treated with less courtesy than other people are.
- You receive poorer service than other people at restaurants or stores.
- People act as if they think you are not smart.
- People act as if they are afraid of you.
- People act as if they think you are dishonest.
- People act as if they’re better than you are.
A follow up question for each item then asked if the reason for the exprience was due to one of the following response options: a. Ancestry or natural origins
b. Gender
c. Race
d. Age
e. Height or weight
f. Some other aspect of physical appearance
g. Never discriminated against
The cronbach's alpha for our modified and shortened scale was 0.81 for discrmination for any reason
- disall: Everyday discrimination score for any reason, counting the number of experiences
- disall_1 through disall_6: refers to individual scale items, discrimination for any reason
- disgender: Everyday discrimination score, counting the number of experiences but only if the reason was due to gender
- disgender_1 through disgender_6: refers to individual scale items, discrimination due to gender only
- disrace: Everyday discrimination score, counting the number of experiences but only if the reason was due to race
- disrace_1 through disgender_6: refers to individual scale items, discrimination due to race only
The cronbach's alpha for our modified and shortened scale was 0.81 for discrmination for any reason; 0.58 for discrimination due to gender, and 0.72 for discrimination due to race.
Adolescent Stress Questionnaire scores
The ASQ scale measures adolescent stress with 27 items grouped in several domains.
Citation: Byrne DG, Davenport SC, Mazanov J. Profiles of adolescent stress: The development of the adolescent stress questionnaire (ASQ). Journal of adolescence. 2007;30(3):393-416. doi:10.1016/j.adolescence.2006.04.004
ASQ_1 Arguments at home.
ASQ_2 Disagreements with your parents.
ASQ_3 Lack of understanding by your parents.
ASQ_4 Following rules at home.
ASQ_5 Living at home.
ASQ_6 Not being taken seriously by your parents.
ASQ_7 Having to study things you don’t understand.
ASQ_8 Teachers expecting too much from you.
ASQ_9 Difficulty with some subjects in school.
ASQ_10 Keeping up with schoolwork.
ASQ_11 Getting along with your boy/girlfriend.
ASQ_12 Breaking up with your boy/girlfriend.
ASQ_13 Making the relationship with your boy/girlfriend work.
ASQ_14 Not having enough time for your boy/girlfriend.
ASQ_15 Pressure to fit in with peers.
ASQ_16 Being hassled for not fitting in.
ASQ_17 Peers hassling you about the way you look.
ASQ_18 Being judged by your friends.
ASQ_19 Concern about your future.
ASQ_20 Having to make decisions about future work or education.
ASQ_21 Putting pressure on yourself to meet your future goals.
ASQ_22 Not getting enough free time.
ASQ_23 Not having enough time for activities outside of school.
ASQ_24 Not having enough time for fun.
ASQ_25 Not having enough money to buy the things you need.
ASQ_26 Not having enough money to buy the things you want.
ASQ_27 Pressure to make more money.
Response options 1=Not at all stressful; 2. A little stressful; 3. Somewhat stressful; 4. Very stressful; 5. Extremely stressful
Scoring Items are summed into each subscale and all items are summed for the total stress scale.
Variables
- stresshomex: ASQ stress at home score-standardized
- stressschx: ASQ stress school performance score-standardized
- stressromx: ASQ stress romantic partner score-standardized
- stresspeerx: ASQ stress peers score-standardized
- stressfutx: ASQ stress future score-standardized
- stressbalancex: ASQ stress school leisure balance score-standardized
- stressmoneyx: ASQ stress money score-standardized
- stresstotalx: ASQ stress total score-standardized
Perceived Stress Scale score
Citation: Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24(4), 385–396.
Items:
PSS_1 In the last month, how often have you been upset because of something that happened unexpectedly?
PSS_2 In the last month, how often have you felt that you were unable to control the important things in your life?
PSS_3 In the last month, how often have you felt nervous and “stressed”?
PSS_4 In the last month, how often have you felt confident about your ability to handle your personal problems?
PSS_5 In the last month, how often have you felt that things were going your way?
PSS_6 In the last month, how often have you found that you could not cope with all the things that you had to do?
PSS_7 In the last month, how often have you been able to control irritations in your life?
PSS_8 In the last month, how often have you felt that you were on top of things?
PSS_9 In the last month, how often have you been angered because of things that were outside of your control?
PSS_10 In the last month, how often have you felt difficulties were piling up so high that you could not overcome them?
Response options 0=Never; 1=Almost never; 2=Sometimes; 3=Fairly often; 4=Very often
Scoring To calculate a total PSS score, responses to the four positively stated items (items 4, 5, 7 and 8) first need to be reversed (i.e. 0 => 4; 1 => 3; 2 => 2; 3 => 1; 4 => 0). The PSS score is then obtained by summing across all items. Higher scores indicate higher levels of perceived stress.
- pssx: Perceived stress scale score-standardized
Code/Software
All code was written for Stata. Each file produces the results for the corresponding tables and figures. Note, some stata code requires the elapse package which tracks the amount of time it takes for a program to run. To install this command, type in the stata command line "findit elapse" and install the elapse package.
File: Infile_riseup-stress.do
Description: Stata code for infiling riseup-stress-public.csv into a Stata format file.
File: Table_1.do
Description: Table 1. Comparison of lottery waitlist and winners, n (%)
File: Table_2.do
Description: Table 2. Study sample age and the standardized mean and range of the stress measures.
File: Table_3_ASQ.do
Description: Table 3. Differences (se) in stress between the lottery winners and waitlist groups among females and males. Analysis for the ASQ (adolescent stress questionnaire) outcomes.
File: Table_3_PSS.do
Description: Table 3. Differences (se) in stress between the lottery winners and waitlist groups among females and males. Analysis for the PSS (perceived stress scale) outcomes.
File: Table_4.do
Description: Table 4. Unadjusted proportion reporting "yes" for each item of the Everyday Discrimination Scale
File: Table_5.do
Description: Everyday Discrimination Scores for the lottery winners and waitlist groups stratified by gender
File: Table_5_sensitivity.do
Description: The sensitivity analysis of Table 5 excludes latino ethnicity, usbirth, native language, parental birthplace in case including those covariates is masking differences in discrimination
File: Figures_1_and_2.do
Description: Creates Figures 1 and 2. Uses output files created in Table_3_ASQ.do and Table_3_PSQ.do files.
File: boots.ado
Description: wrapper code for running bootstrap command and allows for easier implementation of the bootstrap command. Type help boots to get information on using this user written command. This file is usually inserted in the directory for personal ADO files. To determine where to place this file, in the STATA command line type "sysdir". This will list the PERSONAL directory for ADO files. Copy this file to the PERSONAL ADO directory
File: boots.hlp
Description: help file for the boots.do file. This file is usually inserted in the directory for personal ADO files. To determine where to place this file, in the STATA command line type "sysdir". This will list the PERSONAL directory for ADO files. Copy this file to the PERSONAL ADO directory.
Human subjects data
We have obtained IRB approval and consent from subjects to publish de-identified data in the public domain. All participating subjects completed informed assent at study enrollment since subjects were minors at that time and a parent or legal guardian provided written informed consent. Subjects provided written consent to participate in a additional surveys after age 18.
All personal identifiers, including names, IDs, addresses, and other contact information have been removed from the dataset.