Impact of Social Context on the Self-Concept of Gay and Lesbian Youth: A Systematic Review

			Introduction		Methods
Number	Author (Year)	Title/ Abstract	Background Rationale	Objectives	Study Design	Setting	Participants	Variables	Data Sources/ Measurement	Bias	Study Size	Quantitative Variables	Statistical Methods
1	Detrie & Lease (2007)	1	2	2	0	2	2	2	2	1	2	1	1
2	Dahl & Galliher (2010)	1	2	1	0	1	2	0	1	0	1	1	2
3	Grossman & Kerner (1998)	1	2	2	1	2	2	2	2	0	1	2	1
4	Ioverno et al. (2016)	2	2	2	2	2	2	2	2	1	2	2	2
5	Poteat et al. (2015)	1	2	2	1	2	2	1	2	1	2	2	1
6	Hershberger & D’Augelli (1995)	2	2	1	2	2	2	2	2	0	2	2	2
7	Savin- Williams (1989)	1	1	1	1	2	2	2	2	1	2	2	1
8	Snapp et al. (2015)	2	2	2	2	2	2	1	2	0	2	2	1
9	Woodford et al. (2015)	1	2	1	0	1	1	1	2	2	0	2	1
10	Rotheram- Borus et al. (1995)	1	2	1	1	2	2	2	2	2	2	2	1
11	D’Augelli & Hershberger (1993)	1	2	0	0	2	2	1	2	1	2	1	2
12	Russel et al. (2014)	1	2	2	1	1	1	2	2	0	1	2	1
13	Rosario et al. (2005)	2	2	2	1	2	2	1	2	0	2	2	2
14	Blais et al. (2014)	1	2	2	1	2	2	1	2	2	2	2	2
15	Wilkerson et al. (2017)	2	2	1	1	0	1	2	2	1	0	2	2
16	Vincke & van Heeringen (2002)	2	1	2	2	2	2	2	2	2	2	2	2
17	Bauermeister et al. (2010)	2	2	2	2	2	1	2	2	2	2	2	2
18	Rosario et al. (2011)	2	2	2	2	2	2	2	2	2	1	2	1
19	Bos et al. (2008)	2	2	2	2	2	2	1	1	0	2	2	2
20	Toomey et al. (2011)	2	2	2	2	2	2	2	2	2	2	2	2

		Results					Discussion
Number	Author (Year)	Participants	Descriptive Data	Outcome Data	Main Results	Other Analysis	Key Results	Limitations	Interpretation	Generalizability	Funding	Total Score
1	Detrie & Lease (2007)	2	1	1	1	1	2	2	2	2	0	32
2	Dahl & Galliher (2010)	0	2	2	1	2	2	2	2	2	0	27
3	Grossman & Kerner (1998)	0	1	2	2	1	2	1	2	2	0	31
4	Ioverno et al. (2016)	2	2	2	2	2	2	2	2	2	2	43
5	Poteat et al. (2015)	0	1	2	1	2	2	2	2	1	2	34
6	Hershberger & D’Augelli (1995)	1	2	2	1	2	2	2	2	2	2	39
7	Savin- Williams (1989)	0	1	2	1	1	2	2	2	1	0	30
8	Snapp et al. (2015)	0	1	2	1	1	2	2	2	2	2	35
9	Woodford et al. (2015)	0	1	1	1	1	2	2	2	2	2	28
10	Rotheram- Borus et al. (1995)	1	1	2	1	2	2	1	2	2	2	36
11	D’Augelli & Hershberger (1993)	2	1	2	1	2	2	1	1	2	2	32
12	Russel et al. (2014)	0	0	2	2	2	2	2	2	2	0	30
13	Rosario et al. (2005)	0	2	2	1	2	2	2	2	2	2	37
14	Blais et al. (2014)	0	2	2	2	2	2	2	2	2	2	39
15	Wilkerson et al. (2017)	0	2	2	2	2	1	1	1	0	2	29
16	Vincke & van Heeringen (2002)	2	2	2	2	2	2	2	2	2	2	43
17	Bauermeister et al. (2010)	2	2	2	1	1	1	2	2	2	2	40
18	Rosario et al. (2011)	1	1	2	1	2	2	2	2	2	2	39
19	Bos et al. (2008)	0	2	2	1	2	2	2	2	2	2	37
20	Toomey et al. (2011)	0	2	2	2	2	2	2	2	2	2	42

Table 2

Descriptive characteristics of studies

	Author (Year) Country	Design	Sample size (N)/ Age Range/Sexual Orientation	Sexual Orientation Measure	Social Context Measure	Self-Concept Measure	Level of Analysis	Threats
1	Detrie & Lease	CS	N = 218	PIQ	PSS-Fa;	CSES;	Individual	Sample demographics
	(2007)				PSS-Fr;	PWBS (self-	(online survey)	(Caucasian, middle to
			Age = 14–22		SCS;	acceptance)		upper class)
	USA				PWB			Participants already
			L, G, B, ONH					out
								Correlational design
								(no causal inferences)
								Self-report measures
2	Dahl & Galliher	CS	N = 106	Demographic	MOGS;	RSES	Individual	Small sample size
	(2010)			information/sexual	CES-DS;		(online survey)	Convenience sampling
			Age = 18–24	orientation history	RCSS;			Self-report measures
	USA				BMMRS
			L, G, B, ONH
3	Grossman &	CS	N = 90	Sociodemographic	SNS	RSES	Individual	Small sample size
	Kerner (1998)			and risk factor			(in person)	Convenience sample
			Age = 14–21	questionnaire				Geographically limited,
	USA							urban sample
			L, G					Caucasians
								underrepresented
								Population not
								representative
								Self-report measure
								Issues with
								generalizability and
								external validity
4	Ioverno et al.	Long	N = 327	Demographic	GSA presence and	RSES	Individual	Small sample size
	(2016)			information	participation;		(questionnaire—in	Single item measure
			Age = 15–21		Homophobic bullying		person or online)	used for two key
	USA				(single-item question			indicators
			L, G, B, Q		5-point);			Geographically
					School safety (4-point			limited sample (limits
					single question)			generalizability)
5	Poteat et al.	CS	N = 146	Demographic	Victimization (verbal,	RSES	Individual	No causal inferences
	(2015)			information	relational, and physical);		(in person)	Population not
			Age = 14–19		GSA Support and			representative
	USA				Advocacy;			Sample geographically
			L, G, B, Q, ONH		NRI			limited (issues with
								generalizability)
								Small sample size
								Self-report measures
6	Hershberger	CS	N = 165 (22 omitted—	Sampling done is	Victimization variables	RSES;	Individual	Small sample size
	& D’Augelli		unspecified sex,	lesbian and gay	(Attack I, II, III	single-item	(individual—surveys	Sample not
	(1995)		or identified as	community centers	representing escalating	question on	mailed to youth	representative
			heterosexual, or age > 21)		levels of violence—	“Comfort	groups)	Small number of
	USA		(+34 omitted—missing		frequency to be assessed	with sexual		female participants
			data)		0–3);	orientation”		Self-report measures
					family support variables			Interpretation of
			Age = 15–21		(family acceptance 1–4,			developmental
					family protection 0–4,			processes may be
			L, G, B		family relations 0–3);			speculative without
					BSI			longitudinal design
7	Savin-Williams	CS	N = 317	GAL-Q	GAL-Q (parental	RSES	Individual	No causal inferences
	(1989)				knowledge of their		(surveys)	Small number of
			Age = 14–23		child’s homosexuality;			female participants
	USA				satisfaction with maternal			Convenience and
			L, G		and paternal relationship;			snowball sampling
					contact with parents)			(issues with
								generalizability)
								Ethnic minorities
								underrepresented
								Low education
								individuals
8	Snapp et al.	CS	N = 245	Personal	FAS; Retrospective	RSES;	Individual	underrepresented ESM is subjective
	(2015)			characteristics	ESM (friend support;	LGBT-SES		Self-report measures
			Age = 21–25	questionnaire	community support;			Convenience sampling
	USA, Canada				young adult adjustment			Geographically
			L, G, B, T		and well-being; life			limited (issues with
					situation)			generalizability)
								Retrospective cross-
								sectional design
								(recall bias)
								No causal inferences
9	Woodford et al.	CS	N = 187	Two single-item	Microaggressions (distal/	RSES	Individual	No causal inferences
	(2015)			measures on sexual	proximal);		(anonymous web-	Small sample size
			Age = 18–25	orientation and	GAD		based survey)	Whites
	Canada, USA			gender identity/				overrepresented
			L, G, B, Q, ONH	LGBQ identity				Self-report measures.
				salience (one-item				Convenience
				question—4-point				sampling. Internal
				scale)				consistency of self-
								developed scales may
								be questionable
10	Rotheram-	Long	N = 136 (+5	Sampling done at	Gay-related events	RSES	Individual	Small sample size
	Borus et al.		omitted—identified	LGBT drop-in center	(7-item measure			Convenience sample
	(1995)		as heterosexual) (+6		constructed by authors);			Exclusively male
			withdrawn)		LEC			sample (issues with
	USA							generalizability)
			Age = 14–19
			G, B, Q
11	D’Augelli &	CS	N = 194 (+27 omitted)	Questions on sexual	Social Aspects of Sexual	RSES	Individual	Small sample size
	Hershberger			orientation	Orientation (openness;		(questionnaires—	Urban sample
	(1993)		Age = 15–21		conformity; involvement;		filled in individually	Underrepresentation
					disclosure-related		during gay/lesbian	of females (27%).
	USA		L, G, B		events);		group meetings)	Overrepresentation of
					Disclosure of Sexual			whites (66%)
					Orientation Within the			No causal inferences
					Family (items taken from
					Savin-Williams, 1990);
					Mental Health Problems
					(items taken from Mapou
					et al., 1983);
					BSI
	Country		Age Range/Sexual	Measure		Measure
			Orientation
12	Russel et al.	CS	N = 245	Demographic	Self-reported past LGBT	RSES	Individual	Geographically
	(2014)			information taken	school victimization		(computer assisted/	limited (issues with
			Age = 21– 25	from Family	(middle or high)—10 items		paper and pencil)	generalizability)
	USA			Acceptance Project	measured;			Small sample size
			L, G, B, Q	young adult survey	CES-D;			No causal inferences.
					Young adult life			Retrospective
					satisfaction			accounts (recall bias)
								Self-report measures
13	Rosario et al.	Long	N = 156 (+8 omitted)	Baseline interview	Suicidality (questions at	RSES	Individual	Presuicidal
	(2005)			to establish sexual	baseline interview);			psychological distress
			Age = 14–21	orientation	BSI;			not controlled for.
	USA				DSM-III-R (conduct			Retrospective ESM
			L, G, B, ONH		problems);			(recall bias)
					PSS;			Geographically
					SOS;			limited (limited
					MCSDS			generalizability)
								Small sample size.
								Convenience sampling
14	Blais et al.	CS	N = 300	Demographic	Homophobic bullying	SDQ (4 items	Individual	Mainly urban sample.
	(2014)			information	(exclusion and rejection;	used)	(self-report	Underrepresentation
			Age = 14–22		humiliation; damage		questionnaire)	of males
	Canada				to reputation—5-point			Issues with
			L, G, B, Q		nominal scale);			representativeness
					LGBIS (4 items in			and generalizability
					internalized homophobia)			No causal inferences
								Self-report measures
								Physical bullying not
								included
15	Wilkerson et	CS	N = 108	Hatch Youth	Duration of Hatch Youth	One item	Individual	No causal inferences
	al. (2017)			(Houston Area	attendance (1 item);	(5-point scale):	(survey)	Small sample size.
			Age = 13–20	Team Coalition	MSPSS;	“I have high		Convenience sampling
	USA,			of Homosexuals)	CES-DS	self-esteem”
	Netherlands		L, G, B, ONH	recruitment site
16	Vincke & van	Long	N = 197	Demographic	GHQ (5 items from the	RSES	Individual	Small sample size
	Heeringen			information	depression subscale);		(questionnaire)	Convenience sampling
	(2002)		Age = 15–25		BHS;			High attrition rate
					SSS			(36.9%)
	Belgium		L, G					Geographically limited
								Weak
								representativeness
								and generalizability.
								Underrepresentation
								of females
								Self-report measures
17	Bauermeister	Long	N = 350	Kinsey Scale (Kinsey	BDI-II;	RSES	Individual	Convenience sample
	et al. (2010)			et al., 1948)	BSI;		(interviews;	Urban sample
			Age = 15–19		MSPSS;		paper and pencil	No clear breakdown
	USA				ISID		questionnaire)	of sexual orientation
			L, G, B, ONH					of participants. High
								attrition rate (33.7%).
								Self-report measure.
								Geographically limited
								Issues with
								representativeness
								and generalizability
18	Rosario et al.	Long	N = 156 (+8 omitted—did	Demographic	Involvement in LGB-	RSES	Individual	Small sample size
	(2011)		not meet eligibility	information	related activities (28-item		(interview;	Urban sample
			criteria)		checklist; Rosario et al,		questionnaire)	Geographically
	USA		Age = 14–21		2001);			limited (issues with
			L, G, B		NHAI; BSI; PSS-Fa;			generalizability).
					PSS-Fr; SOS; GRSE;			Longitudinal design
					MCSDS			but only short term
								(1 year)
19	Bos et al.	CS	N = 866	Single question on	ADM;	SPPA (Social	Individual	Reliance on
	(2008)			same-sex attraction	EARSI;	Acceptance		self-reports.
			Age = 13–15		MBRS (Respect	subscale);		Geographically
	Netherlands,				subscale);	RSES		limited (issues with
	USA		L, G, B		GHQ (depression);			generalizability)
					ISQ			No causal inferences
20	Toomey et al.	CS	N = 245	Demographic	GSA presence;	RSES	Individual	Small sample size.
	(2011)			information	School victimization (10		(computer assisted/	Geographically
			Age = 21–25		items adapted from the		paper and pencil)	limited (issues with
	USA				California Healthy Kids			generalizability)
			L, G, B, T		Survey, 2010);			No causal inferences.
					CES-D			Retrospective design
								(recall bias)
								Self-report measures

Table 3

Findings of studies

#	Author (Year)	Self-Concept vs. Sexual Identification	Social Context vs. Self-Concept	Social Context vs. Sexual Identification
1	Detrie & Lease (2007)		Positive correlation between social support from friends and self-esteem (r=0.68, p<0.001)
			No correlation between social support from family and self-esteem (r=0.01) Positive correlation between social connectedness and self-esteem (r=0.30, p<0.001)
			Collective self-esteem positively related to self-acceptance (r=0.38, p<0.001) Self-esteem positively related to social support from family (r=0.41, p<0.001) Self-esteem positively related to social support from friends (r=0.41, p<0.001)
2	Dahl & Galliher (2010)	Self-esteem negatively correlated with sexual orientation conflict (r=–0.399, p<0.01)	Self-esteem positively correlated with positive God (r=0.254, p<0.01) and with positive faith (r=0.195, p<0.05) Self-esteem negatively correlated with negative God (r=–0.366, p<0.01) and with fear and guilt (r=–0.404, p<0.01)	Positive correlation between sexual orientation conflict, and negative God (r=0.240, p<0.01) and fear and guilt (r=0.370, p<0.05)
3	Grossman & Kerner (1998)	No significant differences between self-esteem scores for gay males (M=19.7576, n=58, SD=4.92) and lesbians (M=19.00, n=32, SD=5.83)	Higher self-esteem is a moderately strong predictor of lower emotional distress in gay males (R=0.26, F_{(2, 53)}=9.36, p<.001) and a strong predictor in lesbians (R=0.51, F_{(2, 29)}=14.85, p<0.001)	No significant differences in satisfaction with support scores between gay males (M=3.69, n=58, SD=0.53) and lesbians (M=3.66, n=32, SD=0.57) No significant differences in emotional distress scores between gay males (M=1.17, n=58, SD=0.75) and lesbians (M=1.23, n=32, SD=0.79)
4	Ioverno et al. (2016)		GSA presence leads to no changes in self- esteem from T₁ (M=21.14, n=327, SD=5.64) to T₂ (M=20.72, n=327, SD=5.34)	N/A
5	Poteat et al. (2015)	N/A	Victimization negatively correlated with self-esteem (r=–0.18, p<0.5) Higher perception of support from GSA predicted higher scores on self-esteem (β=0.12, p<0.05)	N/A
6	Hershberger & D’Augelli (1995)	Self-esteem positively correlated with sexual identification comfort (r=0.33, p<0.05)	Self-esteem positively correlated with family relations (r=0.11, p<0.05) Self-esteem showed no significant correlation with family acceptance (r=0.03), or with family protection (r=–0.01) Self-esteem negatively correlated with mental health problems (r=–0.46, p<0.05)	Sexual identification comfort positively correlated with family acceptance (r=0.18, p<0.05), with family protection (r=0.21, p<0.05), and with family relations (r=0.44, p<0.05)
7	Savin-Williams (1989)	Self-esteem scores slightly higher for lesbians (M=22.18) than for gay males (M=21.98)	Self-esteem shows no association with parental acceptance, contact with parents, relationship with father, or marital status of parents	Lesbians scored higher (M=2.71) than gay males (M=2.41) on being out to their father
		Self-esteem shows no association with sexual orientation comfort	Lesbians’ self-esteem shows significant correlation with mother’s age (r=–0.23, p<0.02) and satisfaction with relationship with mother (r=–0.26, p<0.01) Gay males’ self-esteem shows significant correlation with being out to their mother (r=–0.15, p<0.03) and with satisfaction with mother (r=–0.22, p<0.003) and with father (r=–0.22, p<0.002)	Lesbians and gay males scored similar (M=2.14 and M=2.15, respectively) on being out to their mother
8	Snapp et al. (2015)	Self-esteem positively correlated with LGBTesteem (r=0.41, p<0.05)	Self-esteem positively correlated with support from friends about being LGBT (r=0.19, p<0.05)	Being out to friends positively correlated with LGBT-esteem (r=0.20, p<0.05) Support from friends about being LGBT
			Self-esteem positively correlated with the availability of LGBT books and magazines (r=0.16, p<0.05) Family acceptance predicted self-esteem (β=0.38, p<0.001) Support from friends about being LGBT predicted self-esteem (β=0.15, p<0.05)	positively correlated with LGBT-esteem (r=0.19, p<0.05) Being out to family, friends, and others predicted LGBT-esteem (β=0.31, p<0.001) Family Acceptance predicted LGBT-Esteem (β=.36, p<.001).
9	Woodford et al. (2015)	LGBQ identity salience positively correlated with self-esteem (r=0.17, p<0.05)	Self-esteem negatively correlated with anxiety (r=–0.55, p≤0.001) Self-esteem negatively correlated with perceived stress (r=–0.71, p≤0.001)	Relationship recognition is negatively correlated with LGBQ identity salience (r=–0.15, p<0.05) Distal environmental microaggressions were positively correlated with LGBQ identity salience (r=0.30, p≤0.001)
10	Rotheram- Borus et al. (1995)	N/A	Self-esteem positively but not significantly correlated with gay-related stressors (r=0.15, p<0.05), with academic stressors (r=0.03, p<0.05), with other life stressors (r=0.15, p<0.05), or with emotional distress (r=0.11, p<0.05)	N/A
11	D’Augelli & Hershberger (1993)	N/A	Self-esteem negatively correlated with mental health issues (r=–0.44, p<01) Self-esteem negatively correlated with depression (r=–0.49, p<0.01) Self-esteem negatively correlated with interpersonal sensitivity (r=–0.41, p<0.01)	N/A
12	Russel et al. (2014)	Self-esteem negatively correlated with hiding at school (r=–0.21, p<0.001) Self-esteem showed no significant correlation with being out in school (r=0.10)	Self-esteem negatively correlated with LGBT victimization (r=–0.24, p<0.05) Self-esteem negatively correlated with depression (r=–0.47, p<0.001) Self-esteem positively correlated with life satisfaction (r=0.34, p<0.001)	Hiding at school was positively correlated with LGBT victimization (r=0.14, p<0.05) Hiding at school positively correlated with depression (r=0.24, p<0.001) Hiding at school negatively correlated with life satisfaction (r=–0.14, p<0.05) Being out at school positively correlated with LGBT victimization (r=0.29, p<0.001) Being out at school negatively correlated with depression (r=–0.22, p<0.001) Being out at school positively correlated with life satisfaction (r=0.16, p<0.05)
13	Rosario et al. (2005)	N/A	At baseline: Self-esteem negatively correlated with depression (r=–0.62, p<0.05), with anxiety (r=–0.35, p<0.05), and with conduct problems (r=–0.16, p<0.05) Self-esteem positively correlated with social support from family (r=0.27, p<0.05) and friends (r=0.28, p<0.05), and with social desirability (r=0.25, p<0.05) Self-esteem negatively correlated with negative social relationships (r=–0.44, p<0.05)	N/A
			At 6-month assessment: Self-esteem negatively correlated with depression (r=–0.54, p<0.05) and with anxiety (r=–0.28, p<0.05). Correlation between self-esteem and conduct problems was negative but not significant (r=–0.10)
			At 12-month assessment: Self-esteem negatively correlated with depression (r=–0.37, p<0.05) and with anxiety (r=–0.23, p<0.05). Correlation between self-esteem and conduct problems was negative but not significant (r=–0.02)
14	Blais et al. (2014)	N/A	Self-esteem negatively correlated with homophobic bullying (r=–0.171, p<0.05) Self-esteem negatively correlated with internalized homophobia (r=–0.171, p<0.05) Internalized homophobia positively correlated with homophobic bullying (r=0.388, p<0.05) Relationship between self-esteem and homophobic bullying mediated by internalized homophobia	N/A
15	Wilkerson et al. (2017)	N/A	Increased social support associated with increased self-esteem (β=–0.72, p<0.05) Hatch Youth attendance was associated with increasing self-esteem (T₁: β=0.08, p<0.05; T₂: β=0.30, p<0.05)	N/A
16	Vincke & van Heeringen (2002)	N/A	Self-esteem positively correlated with parents’ awareness of sexual orientation at T₁ and T₂, but these correlations are not significant (r=0.09 and r=0.09, respectively) Self-esteem negatively correlated with parent acceptance at T₁ (r=–0.03) and positively correlated at T₂ (r=0.08), but these correlations are not significant Self-esteem positively correlated with satisfaction with lesbian/gay friendship relations at T₁ (r=0.20, p<0.05) and at T₂ (r=0.26, p<0.05) Self-esteem positively correlated with confidant support at T₁ (r=0.41, p<0.05) and at T₂ (r=0.41, p<0.05)
17	Bauermeister et al. (2010)	Self-esteem was not associated with Kinsey Scale scores	Males: Self-esteem at Tpositively associated ₂ with self-esteem at T₁ (β=0.47, p≤0.001), with number of friends aware of sexual identity at T(β=0.15, p≤0.05), and with ₂ social support at T₂ (β=0.17, p≤0.05) Females: Self-esteem at Tpositively associated ₂ with self-esteem at T ₁ (β=0.63, p≤0.001) and with social support T ₂ (β=0.20, p≤0.01)	N/A
18	Rosario et al. (2011)		Self-esteem positively associated with social desirability (r=0.28, p<0.05) Self-esteem was positively associated with family support (r=0.26, p<0.05) Self-esteem negatively associated with negative social relationships (r=–0.36, p<0.05)	Highly integrated youth reported less anxiety and depression, and fewer conduct problems
			Highly integrated youth reported higher self-esteem
19	Bos et al. (2008)	Sexual attraction negatively correlated with self-esteem (r=–0.14, p<0.001)	Self-esteem positively correlated with disclosure to father (r=0.26, p<0.001), with disclosure to mother (r=0.18, p<0.001), with social acceptance (r=0.38, p<0.001), and with respect from mentor (r=0.15, p<0.001)	Sexual attraction negatively correlated with disclosure to father (r=–0.10, p<0.01), with social acceptance (r=–0.09, p<0.01), with respect from mentor (r=–0.09, p<0.01), and with school identification (r=–0.09, p<0.01)
			Self-esteem negatively correlated with peer role strain (r=–0.21, p<0.001), and with depression (r=–0.66, p<0.001)	Sexual attraction negatively correlated with disclosure to mother, but correlation was not significant (r=–0.04) Sexual attraction positively correlated with peer role strain (r=0.10, p<0.01) and with depression (r=0.26, p<0.001)
20	Toomey et al. (2011)		Self-esteem positively associated with GSA presence (r=0.12, p<0.01)
			Self-esteem showed no significant association with GSA participation (r=0.06) or with perceived GSA effectiveness (r=0.05)