Dimensionality, cross-validation and reliability of Help-Seeking Dilemma Scale

Abstract

Background

Help-seeking dilemmas are potential barriers to use of treatment, social interventions, and social support systems. While previously existing scales provided a means of assessing help-seeking intentions, a fundamental gap is their limitation to mental health cases or gender lines. Hence, Help-Seeking Dilemma Scale (HSDS) was developed for assessing help-seeking dilemmas regardless of mental health conditions or gender lines.

Methods

The study adopted cross-validation design with 660 participants across three samples. First sample comprised 15 participants (males = 7, females = 8) in an in-depth interview that facilitated the generation of HSDS items. Second sample comprised 446 participants (male = 49.6%, female = 50.0%, unspecified gender = .4%, mean age = 29 ± 11.31 years) randomly split into EFA (n = 227) and CFA (n = 219) subsamples. HSDS was cross-validated with Exploratory Factor Analysis (EFA) in the EFA subsample and Confirmatory Factor Analysis (CFA) in the CFA subsample. Third sample comprised 199 participants (male = 85%, female = 11.5%, unspecified gender = 3.5%, mean age = 31.29 ± 6.55 years) where predictive validity was tested. Data were collected using HSDS, General Help Seeking Questionnaire (GHSQ), Self-stigma of Seeking Help Scale (SSOSH) and Actual Help-Seeking Questionnaire.

Results

EFA revealed sampling adequacy and data appropriateness for factorization [KMO = .910, χ²(231) = 2241.50, p < .0001] and extracted five dimensions, accounting for 62.4% of total variance. CFA confirmed the five-dimensional model of the EFA, with acceptable model fit indices: χ²(160) = 321.041, p < 0.001; CMNI/DF = 2.01; RMSEA = .07 (95% CI: .06 – .08). All HSDS dimensions showed convergent validity. Fornell-Larcker Criterion and Heterotrait-Monotrait (HTMT) ratio confirmed dimensional discriminant validity. Predictive validity showed that HSDS significantly predicted actual help seeking [OR = 1.19, p < 0.001]. Pearson’s correlations confirmed discriminant validity between HSDS and SSOSH, r = .14, p < .01, and GHSQ, r = .05, p > .05. All dimensions demonstrated good internal consistency, with reliability score ranging from .79 to .88. Overall internal consistency for HSDS was .95.

Conclusion

HSDS is valid and reliable for assessing help-seeking dilemmas in academic and general populations and without limitation to any gender.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-026-04690-4.

Introduction

Dilemma refers to uncertainty or being unclear about taking an action [24]. Help-seeking dilemma, therefore, is the uncertainty or being unclear about seeking help despite needing help. The need for help and the specific help needed could be clear to an individual. However, the difficulty is the experience of conflicting motivation, worries, fears that generate some reluctance to seek the needed help. Help-seeking dilemmas could arise from different reasons or difficulties including those perceived or real.

The need to understand and assess help-seeking dilemmas in the general population and specific populations such as clinical and forensic populations has become more important than ever before because of the possible implications and impact it could have on individuals, families, and the society at large. For example, sometimes the failure to seek help could aggravate the possibility of self-harm and fatal suicides [2, 23]. Current data show there is one death by suicide every 43 s with increases of fatal suicide in different regions of the world [13, 22]. Some of these suicide deaths might have been prevented if help was sought.

Why is the assessment of help-seeking dilemma important? First, help-seeking dilemmas are potential psychological barriers to effective treatment outcomes in the management of mental health difficulties. Second, the presence of dilemmas about seeking help could make an individual recline from effective use of social support systems in times of need. Third, the assessment could enable clinicians and professional therapists and counsellors to identify help-seeking dilemmas that are specific to individual clients. Fourth, it could inform professionals about likely behavioural response or adherence of their clients to treatment and how to safeguard against relapse. Fifth, the assessment of help-seeking dilemmas could enable professionals and social service providers in rightly educating families on how to assist family members in distress. Lastly, as mentioned earlier, not seeking help could escalate the possibility of fatal suicides or self-harm. Hence the need to know individuals at risk of not seeking help and the specific dimension of the barrier.

Previous works on the development of help-seeking measures include the General Help-Seeking Questionnaire (GHSQ) by Rickwood et al. [18], Mental Help-Seeking Intention Scale (MHSIS) by Hammer & Spiker [10], Attitudes toward Seeking Professional Psychological Help Scale (ATSPPH) by Fischer & Farina [8], Barriers to Help‑Seeking Scale (BHSS) by Mansfield et al. [16] and Self-Stigma of Seeking Help Scale (SSOSH) developed by Vogel et al. [25] to measure perceived self-stigma in seeking psychological help or therapy. Fundamentally, GHSQ assesses self-rated likelihood of seeking help from different sources such as family, friends and professionals. While MHSIS assesses intention to seek help from mental health professionals, ATSPPH assesses attitude towards seeking psychological help. BHSS assesses perceived barriers to seeking help among men, which limits its wider application to other gender lines.

The limitation of MHSIS, ATSPPH and BHSS is that they assess seeking help for mental health, which could make it difficult to use them in situations not related to mental health difficulties. While GHSQ assesses the likelihood or intention to seek help, it does not identify reasons that could affect such likelihood. SSOSH is a unidimensional scale that measures only self-stigma and ignores other barriers in help-seeking. Meanwhile, help-seeking barriers are multifaceted.

Given the aforementioned limitations of previously existing measures, the present study presents a new tool – Help-Seeking Dilemma Scale (HSDS) – for assessing help-seeking dilemmas that are not captured or identified in any other tool on help seeking. HSDS is intended to assess dilemmas individuals experience when the need to seek help arises. This goes beyond just the intention to seek help. It is specifically about those internalised issues that hinder the intention to seek help. This is important for clinicians, therapists, social service providers and families, because it will enable them identify specific difficulties that hold people back from seeking help.

Additionally, HSDS is designed for use regardless of the context in which the individual is seeking help or the gender of the help seeker. These are what make HSDS unique when compared to others. In the light of this, the aim of this study is to develop a new measure that assesses help-seeking dilemmas, identify its dimensionality and establish its validity and reliability.

Methods

Design

This study used cross-validation design with a survey method for data collection. There were three samples. The first sample involved in-depth interviews (IDI). The second sample was randomly split into two subsamples: EFA and CFA. The EFA subsample was used for Exploratory Factor Analysis while the CFA subsample was used for Confirmatory Factor Analysis to cross-validate HSDS. The third sample was used to test predictive validity of HSDS. HSDS was measured as a continuous variable.

Participants

There were 660 participants across three samples in this study. The first sample comprised 15 participants (males = 7 and females = 8) who voluntarily participated in an in-depth interviews (IDI), whose data were used to facilitate the generation of items for HSDS. Seven experts with a minimum of Master’s degree in Sociology, reviewed the items generated for content validation. The second sample comprised 446 students (undergraduates and postgraduates) and staff (academic and non-academic) of the University of Ibadan, Oyo state, Nigeria. This institution deemed fit because it is a federal university with students from various parts of Nigeria and international students.

The estimated population size was 40,000 students and 5,000 staff obtained from the website of the University of Ibadan. Using Slovin’s formula: N/(1 + Ne²) where population size estimate (N = 45,000) and error margin (e) = 0.05, the formula gives 45,000/1 + (45,000 × 0.05²) = 400. A possible 10% attrition was added 450.

The third sample comprised 199 participants selected from the general population (artisans, traders, etc.) in Ibadan North metropolis, Oyo State, Nigeria. This sample was used to establish predictive validity for HSDS. Due to unknown population size estimate of the people in Ibadan North metropolis, GPower 3.1 was used to estimate the minimum number of participants needed using the following parameters: test = linear regression with R² deviation from zero, effect size = 0.15, power (1 – β) = 0.99, error margin = 0.05 and one predictor (that is, HSDS). This yielded minimum estimate of 125 participants. Eventually, 199 of 200 participants were recruited because one questionnaire was unusable. Bringing all samples together, the overall sample size of this study was 660.

Accidental sampling was employed in this study because of practical difficulty of getting the participants with a random selection method. Additionally, due to the large size and dispersed nature of the population, it was only practically possible to approach each person and invite them to participate in the study. All participants met the inclusion criteria: (1) minimum of 18 years of age (2) consent to participate in the study.

Procedure

The researcher conducted in-depth interviews (IDI) with 15 voluntarily participating individuals, who formed the first sample. Participants were asked to express their thoughts about seeking help generally and to identify the dilemmas they face when they have to seek help. This resulted in the collection of qualitative data reflecting diverse dilemmas of help seeking. With content analysis, participants’ responses were used to generate a pool of 24 statements that captured the dilemmas expressed. Seven Sociologists reviewed the items and requested that two should be removed. The remaining 22 items were administered to participants of the second and third samples together with other items from standardized behavioral scales as shown in the instrument section.

Participants were given an informed consent form which details the title and the objectives of this study to obtain their consent. Participants were able to choose whether to participate or not. A participant could decline their participation without any consequence. There was no financial inducement. Participants were informed that data collected would be used for academic purposes including publication. There was no identifying information requested from the participants.

Instruments

With a structured questionnaire, the authors obtained participants’ demographic characteristics: sex, age (as of last birthday), marital status, religion, and ethnicity. Others scales in the questionnaire are as follows:

Help-Seeking Dilemma Scale (HSDS)

The Help-Seeking Dilemma Scale (HSDS) was developed by the authors. Participants responded to each item by using a 5-point Likert (0 = Never, 1 = Rarely, 2 = Sometimes, 3 = Often, 4 = Always). Items 2, 5, 7 and 9 were reverse coded. Composite scores are obtained by summing all responses. High scores indicate high level of help-seeking dilemma. The dimensions, validity and reliability of HSDS are shown in the results section. A sample item is “I worry I might become vulnerable by seeking help.”

Self-stigma of Seeking Help Scale (SSOSH)

SSOSH is a 10-item scale developed by Vogel et al. [25] to measure perceived self-stigma in seeking psychological help or therapy. It utilizes a 5-point Likert (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree). Items 2 and 9 are reverse-scored. Total score was calculated by summing the responses with high scores indicating high level of perceived self-stigma in seeking psychological help. The authors of SSOSH reported good internal consistency (Cronbach’s alpha = 0.91). A sample item is “My self-confidence would not be threatened if I sought professional help.”

General Help-Seeking Questionnaire (GHSQ)

Wilson et al. [26] developed the GHSQ to assess the likelihood of seeking help from various sources. GHSQ uses odd-numbered Likert (1 = Extremely Unlikely, 3 = Unlikely, 5 = Likely to 7 = Extremely Likely) to rate the likelihood of seeking help among eight professional (e.g., general practitioner or mental health professional) and lay (e.g., parents and friends) possible sources. High scores indicate high likelihood of seeking help in general. The authors of GHSQ reported internal consistency with Cronbach’s alpha = 0.85. Participants are prompted with an anchor statement: “How likely is it that you would seek help from the following people?” Sources included, but not limited to intimate partners, friends, and parents.

Actual Help-Seeking Questionnaire (AHSQ)

Rickwood et al. [18] developed the AHSQ to measure actual help-seeking behaviour with 10 items that reflect different sources of help. Participants responded with “Yes” or “No” to indicate whether they sough help or not in the last 30 days. Binary scoring was used, where 1 = Yes, sought help and 0 = No, did not seek help. High scores indicate more actual help-seeking. Test–retest of AHSQ demonstrated good stability over time frames [18]. Participants are prompted with an anchor statement: “In the past 30 days, have you sought help for a personal or emotional problem from any of the following sources?" There are 10 sources of help, which include friends, intimate partners, parents, professionals, etc.

Data analyses

Across the three samples, data on demographic characteristics were analyzed using descriptive statistics (mean, standard deviation, frequency and percentage). In the first sample, qualitative data obtained from the IDI were used to generate items for HSDS by using content analysis. In other words, items for HSDS were extracted from participants’ responses. This was the primary aim of the IDI. In order to achieve cross-validation of HSDS, data from the second sample were randomly split into two subsamples by using SPSS 27 command to split the dataset into approximately 50% halves. The first half was named EFA subsample, which was used to conduct Exploratory Factor Analysis (EFA) that identified the underlying factor structure. Factor membership of all items was determined by using Varimax Rotation in order to minimise overlapping of items across different factors. Only items with the minimum factor loading (eigenvalue = 0.5) were retained after Varimax Rotation.

The second half was named CFA subsample, which was used to conduct Confirmatory Factor Analysis (CFA) that confirmed the fit between the factor structure extracted with EFA and the dataset from the CFA sample. Average Variance Extracted (AVE) was used to test convergent validity. Fornell-Larcker Criterion and Heterotrait-Monotrait (HTMT) ratio were used to test discriminant validity. Composite reliability (CR) test was used to test internal consistency of HSDS items in both EFA and CFA samples. Pearson’s Correlations was used to examine the relationships between HSDS Scale and GHSQ and SSOSH.

Using scores from AHSQ, the researchers classified participants who had sought help from at least one (1) source in the last 30 days into one group (1 = Sought-help) and placed those who did not seek help from any source in the past 30 days into another group (0 = Did not seek help). In order to achieve predictive validity, Logistics regression tested the ability of HSDS to predict actual help seeking behaviour between both groups (Sought help vs Did not seek help). Statistical Package for the Social Sciences (SPSS) 27 was used to conduct EFA and regression analyses. AMOS 22 was used to conduct CFA. Microsoft Excel 2016 was used to conduct AVE, HTMT and CR.

Results

Descriptive analysis of demographic distributions is shown in Table 1 as follows:

Table 1.

Demographic distributions

Characteristics	Frequency (n = 646)	Percent	Mean	Std. Dev
Age			29.73	10.12
Gender
Male	390	60.5
Female	246	38.1
Prefer not to say	9	1.4
Marital Status
Never married	402	62.3
Married	217	33.6
Previously married	26	4.0
Religion
Christianity	476	73.8
Islam	151	23.4
Others	18	2.8
Ethnicity
Yoruba	505	78.3
Igbo	85	13.2
Hausa	4	.6
Others	51	7.9

Table 1 shows the demographic distribution of the 646 participants (excluding the 15 persons who participated in the IDI), which comprised 390 (60.5%) males, 246 (38.1%) females, and 9 (1.4%) participants who preferred not to identify their gender. The average age was 29.73 (SD = 10.12) years. There were 402 (62.3%) unmarried, 217 (33.6%) married and 26 (4.0%) who were previously married. Most of the participants were Christians (73.8%) and Yorubas (78.3%).

Exploratory factor analyses of HSDS in the EFA sample

Using Exploratory Factor Analyses (EFA), HSDS scores in the EFA sample were analysed for factor structure. To determine the suitability of the data for factor analysis, the Kaiser–Meyer–Olkin (KMO) Measure of Sampling Adequacy and Bartlett’s Test of Sphericity were conducted. The KMO and Bartlett’s tests showed excellent sampling adequacy and appropriateness of the data for factor extraction [KMO = 0.910, χ² (231) = 2241.50, p < 0.0001]. The Principal Component Analysis (PCA) revealed five factors based on eigenvalues greater than 1, which accounted for 62.4% of the total variance. In order to achieve a more precise item membership of the factors by minimizing cross-loading of items, orthogonal rotation with varimax method was applied. Additionally, items were retained based on two criteria: (1) a minimum factor loading of 0.50, and (2) conceptual consistency with other items within the factor on which the item demonstrated its highest loading. Based on these criteria, two items (11 and 4): “I worry that seeking help might compromise my privacy” and “I fear feeling embarrassed or ashamed when I ask for help”, were removed, leaving 20 items. The result is shown in Table 2.

Table 2.

Extracted factors and their item loadings

S/N	Items	Factor Loadings
		1	2	3	4	5
8	I fear being labelled by my problem if I seek help	.758
6	I fear being blamed for my problems if I ask for help	.740
9	I fear being rejected when I ask for help	.689
7	I worry about receiving dismissive responses when I seek help	.618
5	I worry that seeking help will cause others to perceive me negatively	.615
10	I feel guilty when I ask for help	.542
20	I seek help only in desperate situations		.713
21	I struggle with opening up to others when I need help		.644
19	I feel reluctant to ask for help		.635
15	I feel like I am burdening others when I ask for help		.633
22	I feel confused about seeking help from others		.537
*11	I worry that seeking help might compromise my privacy		.449
1	I fear the possibility of a trust violation if I seek help			.814
2	I worry I might become vulnerable by seeking help			.710
3	I worry about being betrayed if I seek help			.594
*4	I fear feeling embarrassed or ashamed when I ask for help			.476
17	I find it difficult to seek help from the opposite gender				.813
18	Seeking help from the opposite gender causes me to feel insecure				.796
16	Seeking help makes me uncomfortable because of my gender				.667
13	I believe that people are generally insincere when giving help					.822
14	I believe that people rarely genuinely care about others when offering help					.714
12	I perceive that people are socially apathetic towards helping others					.561

^*Items removed

Table 2 shows five factors were extracted which represent the dimensions of HSDS. Based on the thematic grouping of the items, it shows the first factor comprises items that assess help-seeking social rejection. The second factor shows items that assess help-seeking avoidance. The third factor comprises items that assess help-seeking vulnerability. The fourth factor comprises items that assess gendered help-seeking barrier. The fifth and last factor shows items that assess help-seeking cynicism. Consequently, factors 1 to 5 were dilemmas in the form of social rejection (SR), avoidance (A), vulnerability (V), gender barrier (GB) and cynicism (C) respectively.

Confirmatory factor analyses of HSDS in the CFA sample

In order to confirm the factor structure identified in the EFA sample, the factor structure was subjected to CFA using the data from the CFA sample. The result is shown in Fig. 1 as follows.

Fig. 1.

Confirmatory Factor Structure of HSDS Note: SR = Social rejection, A = Avoidance, V = Vulnerability, GI = Gender barrier, C = Cynicism

The confirmatory factory analysis of HSDS in Fig. 1 shows the latent structures: social rejection, avoidance, vulnerability, gender barrier, cynicism about help and their observed variables (items of the scale). All items loaded significantly on their respective dimensions, with standardised loadings ranging from 0.57 to 0.90. This confirms that they are valid. Additionally, there was good inter-factor correlations of estimates as shown along the double-headed arrows.

Based on guidelines from Hooper et al. [12], the model fit indices met the recommended cut-offs, which are indicative of good fit between the data and the five dimensional model suggested in the EFA sample: χ2 (160) = 321.041, p < 0.001; CMNI/DF = 2.01; GFI/AGFI = 0.87/0.83; RMSEA = 0.07 (95% CI: 0.057–0.79); AIC = 421.041; SRMR = 0.08; N = 210. The incremental fit indices, NFI = 0.87, RFI = 0.84, IFI = 0.93, TLI = 0.92, and CFI = 0.92, met acceptable cut-offs [12]. The parsimonious-adjusted indices showed PRATIO = 0.84, PNFI = 0.73 and PCFI = 0.78. Overall, the confirmatory factor analysis from the CFA sample confirmed the hypothesised five-factor model found in the EFA sample.

Convergent, discriminant, and predictive validity of HSDS

Furthermore, the five dimensions (factors) of HSDS was tested for convergent and discriminant validity. First, convergent validity was tested using the Average Variance Extracted (AVE) formula, that is, the sum of all squared standardised loadings, Σ(λ²), divided by the number of indicators in the factor, n. Figure 2 shows how the formula was applied to each dimension, using the first dimension (social rejection) as an example. Social rejection has 6 factor loadings.

Fig. 2.

Average variance extracted

In order to determine the internal consistency of each factor of HSDS, composite reliability (CR) formula was applied, where (Σλ)² is the square of the sum of standardised loadings and Σθ is the sum of error variances. Figure 3 shows how the formula was applied for all HSDS dimensions, using social rejection (SR) as an example.

Fig. 3.

Composite reliability for social rejection

The full results of the average variance extracted and composite reliability scores for each factor is presented in Table 3.

Table 3.

Average variance extracted and composite reliability of HSDS dimensions

Dimensions	Items	Factor Loadings	AVE	SQRT(AVE)	CR
Social rejection	5, 6, 7, 8, 9, 10	.78,.66,.76,.80,.71,.72	0.55	0.74	0.88
Avoidance	15, 19, 20, 21, 22	.57,.80,.80,.75,.73	0.54	0.73	0.85
Vulnerability	1, 2, 3	.80,.77,.76	0.60	0.77	0.82
Gender barrier	16, 17, 18	.87,.90,.61	0.46	0.68	0.79
Cynicism about help	12, 13, 14	.65,.65,.73	0.65	0.81	0.80

AVE Average Variance Extracted, SQRT Square root, CR Composite Reliability

Table 3 shows that convergent validity was supported for all the dimensions, with AVE values being greater than 0.50 recommended threshold (social rejection = 0.55, avoidance = 0.54, vulnerability = 0.60, gender barrier = 0.46, cynicism about help = 0.65) except gender barrier. This implies that, on the average, more than half of the variance in the items is explained by the dimension they represent, except gender barrier that shows 46% variance explanation. Gender barrier was retained because previous findings from empirical reviews indicated that gender differences including transgender and being non-binary, could impact seeking help because of possible discrimination [11]. Additionally, gender barrier demonstrated good internal consistency and discriminant validity as other dimensions.

In order to strengthen validity check of how distinctly different the dimensions are, the authors conducted discriminant validity with Fornell-Larcker Criterion (that is the square root of AVE) and the result (see Table 3) supported discriminant validity of the dimension, with the square root of AVE for each dimension exceeding its correlations with other dimensions. Additionally, Heterotrait-Monotrait (HTMT) ratio was applied in order to support the Fornell-Larcker Criterion and the results (see Table 4) further confirmed discriminant validity among the dimensions, with all HTMT ratios, ranging from 0.68 to 0.81, below the recommended threshold of 0.85, except social rejection –avoidance pair that was at borderline of validity. Additionally, all dimensions demonstrated good internal consistency, with CR score ranging from 0.79 to 0.88. Overall internal consistency for HSDS was 0.95.

Table 4.

Heterotrait-Monotrait discriminant validity

Dimension pairs	Heterotrait average	Monotrait average	HTMT ratio
Social rejection – Avoidance	.41	.48	.85*
Social rejection – Vulnerability	.39	.53	.74**
Social rejection – Gender barrier	.37	.53	.70**
Social rejection – Cynicism about help	.33	.46	.72**
Avoidance – Vulnerability	.42	.52	.81**
Avoidance – Gender barrier	.40	.52	.77**
Avoidance – Cynicism	.36	.45	.80**
Vulnerability – Gender barrier	.38	.56	.68**
Vulnerability – Cynicism about help	.34	.49	.69**
Gender barrier – Cynicism	.35	.50	.70**

^*Borderline valid, ** Valid

Overall, the results show 20 valid items with five dimensions in the HSDS. The five dimensional model was supported by the CFA. All dimensions demonstrated convergent and divergent validities using the AVE, Fornell-Larcker Criterion rule and HTMT. Each of the dimensions and the overall of HSDS have good internal consistency.

In order to enhance the utility of HSDS, its scores were used to predict actual help seeking behaviour scores from AHSQ. The result is shown in Table 5.

Table 5.

Predictive validity of HSDS

Predictor	B	S.E	Wald	Df	Sig	Exp(B)	95% C.I
							Lower	Upper
HSDS	.174	.030	34.173	1	.000	1.190	1.122	1.261

−2Log likelihood (−2LL) = 100.87, Nagelkerke R² =.394, Omnibus test: χ² (1) = 45.90, p <.001, N = 199

Table 5 shows clearly that HSDS scores significantly predicted scores from AHSQ [OR = 1.19, Wald (1) = 34.17, p < 0.001], with HSDS scores predicting minimal increase (19%) in AHSQ scores. This shows that help-seeking dilemma predicts minimal increase of actual help-seeking behaviour. The model statistic shows good fit for the model and HSDS accounts for 39.4% of the total variance of actual help seeking measured with AHSQ [−2LL = 100.87, R² = 0.394, χ² (1) = 45.90, p < 0.001].

Correlations of HSDS with GHSQ and SSOSH

HSDS was designed to measure help-seeking dilemmas that create conflicting motivation with seeking help without limitation to any context or gender of the help seeker. Two of the previously reviewed scales (GHSQ and SSOSH) were chosen to test discriminant validity of HSDS using Pearson’s correlation. These tests were chosen because they are not limited to any gender, neither were they designed for mental health situation exclusive like others. The results of the correlation are shown in Table 6.

Table 6.

Correlation of HSDS and its dimension with SSOSH and GHSQ

Constructs	SSOSH	GHSQ	SR	A	V	GB	C	Mean	SD
Self-stigma of Seeking Help (SSOSH)	–							25.02	6.14
General Help Seeking Questionnaire (GHSQ)	-.249**	–						37.96	8.60
Social rejection (SR)	.166**	.060	–					10.33	5.60
Avoidance (A)	.057	.000	.634**	–				11.10	4.65
Vulnerability (V)	.098*	.042	.596**	.531**	–			5.70	2.79
Gender barrier (GB)	.129**	-.008	.513**	.583**	.375**	–		5.47	3.38
Cynicism about help (C)	.083	.053	.554**	.684**	.436**	.483**	–	5.81	2.57
Help-Seeking Dilemma Scale (HSDS)	.135**	.047	.879**	.859**	.718**	.728**	.751**	37.94	15.00

^**. Correlation is significant at the 0.01 level (2-tailed)

^*. Correlation is significant at the 0.05 level (2-tailed)

Table 6 results show that, first, all dimensions of HSDS correlated very weakly with SSOSH. Overall, HSDS also showed very weak significant correlation with SSOSH, r = 0.14, p < 0.01 and non-significant correlation with GHSQ, r = 0.05, p > 0.05. Additionally, all the dimensions as well as the overall of HSDS did not have any significant correlation with GHSQ. These suggest that HSDS measures a different dimension of help-seeking than GHSQ and SSOSH.

Discussion

This study has provided a valid and reliable measure of help-seeking dilemmas that sometimes characterize people’s disposition toward seeking help. While most of the measures that existed before HSDS provided narrowed help-seeking scope to mental health, the present study provided an assessment that captures a broader scope that is not limited to any situation, health condition or source of help-seeking. Additionally, beyond just help-seeking intention, HSDS captures experiential dilemmas and how frequently they occur with each person. With several layers of validity tests, HSDS demonstrated psychometric validity in capturing what it was designed to capture. Individual dimensions and the overall HSDS items demonstrated strong internal consistency.

The dimensions of HSDS present important focal points in clinical and social interventions. Help-seeking dilemma in the form of anticipation of social rejection could increase the chances that an individual, who is desperate for help, will take irrational decisions such as committing suicide. This is evident in reports where anticipatory social rejection or actual experience of shame and humiliation showed associations with self-harm and suicidality [7, 19]. It is important to re-echo the conclusions of Sadath et al. [19] that the assessment of shame, humiliation, and, of course, anticipatory social rejection, has been a challenge in clinical settings. This lends credence and support to the need for the development of HSDS at this time.

Similar outcomes showed that social rejection increases the likelihood of suicidal ideation and attempts, especially among vulnerable or marginalised populations including adolescents, lesbians and gays, or those who identify as bisexual sexuality or transgender [6, 19]. Given these findings, HSDS is therefore a useful tool which could enable clinical and social intervention service providers estimate the likelihood of seeking help among their clients.

Avoidance dilemma is another important focal point in the risk assessment of individuals seeking help, especially those in populations often neglected such as lesbians, gays, transgender and those with bisexual orientation. In HSDS, help-seeking avoidance implies the reluctance and confusion that shroud the motivation to seek help when it is needed. Previous studies showed that such avoidance has the potential to block the effectiveness of even psychological interventions including the widely used cognitive behavioural therapy [15]. Avoidance could be maladaptive because it weakens the motivation to seek help, which in turn could worsen the situation especially when mental health difficulties are involved [14]. HSDS therefore serves as a measure that could provide insights into the presence of help-seeking avoidance in clients.

The assessment of perception of vulnerability in seeking help is another crucial area that must not be ignored. Sometimes, receiving help comes with the cost of revealing private information that will not be known if the need to ask for help never arises. Having to divulge private information against one’s will creates fear of vulnerability or being at the mercy of others. It also creates the feeling of embarrassment, which has been reported as a predictor of lower likelihood of seeking help [21]. Qualitative studies revealed that feeling of vulnerability as well as fear of judgment reduces the willingness to seek help with depression and social services [3, 9].

Gender barrier in HSDS is the conflicting motivation to seek help from an opposite gender. This has been found to play a role in people’s response to health screenings [17]. The unwillingness to seek help from opposite gender could arise when there is the perception of insecurity or the potential of being taken advantage of. It is also more likely in cultural and religious situations that are more restrictive on women [17]. It is important to note that this is not restricted to women. Gender barrier in HSDS is not particular about any gender direction. Men can equally feel unwilling to seek help from women due to gender differences. Preferences for gender-sameness could pose a barrier to seeking professional help [1]. In cultural settings where conformity to masculine norm is prevalent, men are less likely to seek professional help for their mental health challenges [5]. Assessing gender barriers among patients and clients is therefore important for professionals in order to address such barriers.

Cynicism about true altruism is one of the core dimensions of HSDS. This indicates the extent to which an individual believes people genuinely want to help without an ulterior motive. Being cynical about help could make people undermine trust in others and discourage reliance on social or institutional support [20]. A positive association between cynicism and avoidance-oriented thinking styles could reduce openness to corrective feedback and professional help-seeking [4]. Hence, cynicism is a potential barrier to seeking professional or social service, which can be assessed with HSDS.

Overall, help seeking dilemma scale is a significant predictor of actual help seeking behaviour. The statistical indicators showed that HSDS predicts minimal increases in actual help seeking measured by ASHQ.

Recommendations

The assessment of help-seeking dilemmas should be an integral part of clinical assessments and social services delivery. With HSDS, help-seeking dilemmas can reliably be assessed and identified. This could assist in the development of individualized management plan as much as needed. HSDS can be used among students or general population. Professionals, researchers, policy makers, students and individuals can use it for their research or policy-making purposes.

Implications

This study presents an easy-to-use assessment tool for professionals and social service providers whose activities include providing help to individuals. They can use HSDS to assess dilemmas that limit patients’ service engagement. The outcome of such assessment could inform the design of interventions. Researchers and educationist can use HSDS to assess help seeking dilemmas among students and general population.

Strengths and limitations of the study

A primary strength of this study is the presentation of a tool that assesses help-seeking dilemmas that is not limited by context or gender of the help seeker. Additionally, the items presented assess actual experiences of dilemmas that create motivational conflicts when people attempt to seek help rather than preferences, projected attitude, intention, or beliefs as predominant in previously existing measures on help seeking. This could enable professionals, researchers, and social service providers identify experiential dilemmatic barriers that hold people back from seeking help and build a more tailored intervention that addresses them.

The multiple layers of validation, including cross-validation two independent samples of university students and staff and predictive validity with another sample drawn from the general population, reduce the risk of the HSDS from being sample-specific, enhances generalisability of the results, minimises bias in the validity and reliability qualities of the instrument, and strengthens external validity. Additionally, HSDS does not require any technical training before it could be used. Items are easy to understand and scoring is not complex.

However, HSDS cannot be used as a screening tool. There are no specificity or sensitivity indicators. It does not provide a means of identifying inaccurate responses arising from social desirability and other participant-related biases.

Suggestions for further study

While the present study used cross-validation approach, the authors suggest that future researchers could complement this with other approaches such as the use of Photovoice (PV) and Interpretative Phenomenological Analysis (IPA) for qualitative data analysis to uncover the dimensions of help-seeking and Community-Based Participatory Research (CBPR) in order to obtain participants’ unique experiences in help seeking, enhance sampling adequacy and generalisability of help-seeking instruments.

Conclusions

Help-Seeking Dilemma Scale is a valid and reliable instrument for assessing help-seeking dilemmas among university students and staff and the general population without regard to the gender of the individual. This is relevant for use among professionals, researchers, and social service providers.

Abbreviations

HSDS

Help-Seeking Dilemma Scale

GHSQ

General Help-Seeking Questionnaire

SSOSH

Self-stigma of Seeking Help Scale

EFA

Exploratory Factor Analysis

CFA

Confirmatory Factor Analysis

KMO

Kaiser–Meyer–Olkin

CMNI/DF

Chi-square Minimum/Degrees of Freedom

RMSEA

Root Mean Square Error of Approximation

SRMR

Standardized Root Mean Square Residual

NFI

Normed Fit Index

TLI

Tucker-Lewis Index

CFI

Comparative Fit Index

GFI

Goodness-of-Fit Index

AGFI

Adjusted Goodness-of-Fit Index

AIC

Akaike Information Criterion

RFI

Relative Fit Index

IFI

Incremental Fit Index

TLI

Tucker-Lewis Index

PNFI

Parsimony Normed Fit Index

PCFI

Parsimony Comparative Fit Index

SQRT

Square root

HTMT

Heterotrait-Monotrait

CR

Composite Reliability

PCA

Principle Component Analysis

MHSIS

Mental Help-Seeking Intention Scale

ATSPPH

Attitudes toward Seeking Professional Psychological Help Scale

AHSQ

Actual Help-seeking Questionnaire

-2LL

-2Log likelihood

BHSS

Barriers to Help‑Seeking

SSOSH

Self-Stigma of Seeking Help Scale

IDI

In-depth interviews

SPSS

Statistical Package for the Social Sciences

AMOS

Analysis of Moment Structures

AVE

Average Variance Extracted

SSHREC

Social Sciences and Humanities Research Ethics Committee

SR

Social rejection

A

Avoidance

V

Vulnerability

GB

Gender barrier

C

Cynicism

Authors’ contributions

All authors contributed significantly to the research and endorsed the final manuscript.

Funding

There was no funding received for the execution of this research.

Data availability

De-identified materials and data from this study will be made available by the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The authors adhered to ethical guidelines provided by the Social Sciences and Humanities Research Ethics Committee (SSHREC), University of Ibadan, Ibadan, Nigeria, and the Declaration of Helsinki with respect to human research including the protection of participants’ rights, safety, and well-being ensure the rights, safety, and well-being. SSHREC reviewed the protocol and gave approval for the study with the approval number UI/SSHREC/2025/212. Prior to participation, participants were informed about the study including the title, objectives of the study, how to participate, what information will or will not be requested from them and the use of the data collected for research purposes only including publication. Participants provided written consent. Participation was entirely voluntary and any declined participation was without any consequence. There was no financial inducement. No identifying information was requested from participants.

Consent for publication

Participants were informed that the data collected will be used for academic purposes including publication. No identifying information was requested. Data were entirely de-identified.

Competing interests

The authors declare no competing interests.