
Vol. 12/ Núm. 3 2025 pág. 2506
https://doi.org/10.69639/arandu.v12i3.1489
Discrimination in dating apps in the LGBT community:
Construction and validation of an instrument
Discriminación en aplicaciones de citas en la comunidad LGBT: Construcción y
validación de un instrumento
Julia María Marroquín Figueroa
julia.marroquin@unach.mx
https://orcid.org/0000-0002-2263-3365
Universidad Autónoma de Chiapas
Jorge Alberto Esponda Pérez
jorge.esponda@unicach.mx
https://orcid.org/0000-0002-6821-5361
Universidad de Ciencias y Artes de Chiapas
Mónica Juárez Ibarias
monica.juarez@unach.mx
https://orcid.org/0009-0007-4561-7396
Universidad Autónoma de Chiapas
Juan Manuel Tapia de Aquino
juan.tapia@unach.mx
https://orcid.org/0009-0005-3290-8878
Universidad Autónoma de Chiapas
Dolores Guadalupe Sosa Zúñiga
dolores.sosa@unicach.mx
https://orcid.org/0000-0003-0395-3766
Universidad de Ciencias y Artes de Chiapas
Artículo recibido: 18 julio 2025 - Aceptado para publicación: 28 agosto 2025
Conflictos de intereses: Ninguno que declarar.
ABSTRACT
This study analyzes the Discrimination Scale in LGBT dating apps in adults in Chiapas, Mexico.
In this exploratory and psychometric research, 640 members of the LGBT community actively
participated. A two-factor structure was validated: individual and group discrimination. When
analyzing the frequency of responses for the 17 items, it was observed that items 14 to 16
predominate with responses of “Almost every day”, while for the others, the most common
response was “Every day”. In addition, verifying the Cronbach's Alpha values, it is observed that
the standardized coefficient is .921, the 95% confidence interval is within the range considered
acceptable. These results highlight the ability of the scale to measure this phenomenon in the
LGBT community, suggesting its usefulness both in Chiapas, Mexico, and possibly in other
similar contexts.
Keywords: discrimination, LGBT, dating apps, psychometric properties, validation

Vol. 12/ Núm. 3 2025 pág. 2507
RESUMEN
Este estudio analiza la Escala de Discriminación en aplicaciones de citas LGBT en adultos en
Chiapas, México. En esta investigación exploratoria y psicométrica, participaron activamente 640
miembros de la comunidad LGBT. Se validó una estructura de dos factores: discriminación
individual y discriminación grupal. Al analizar la frecuencia de respuestas para los 17 ítems, se
observó que los ítems 14 a 16 predominan con respuestas de "Casi todos los días", mientras que
para los demás, la respuesta más común fue "Todos los días". Además, al verificar los valores del
Alfa de Cronbach, se observa que el coeficiente estandarizado es de .921, y el intervalo de
confianza del 95% se encuentra dentro del rango considerado aceptable. Estos resultados resaltan
la capacidad de la escala para medir este fenómeno en la comunidad LGBT, sugiriendo su utilidad
tanto en Chiapas, México, como posiblemente en otros contextos similares.
Palabras clave: discriminación, LGBT, aplicaciones de citas, propiedades psicométricas,
validación
Todo el contenido de la Revista Científica Internacional Arandu UTIC publicado en este sitio está disponible bajo
licencia Creative Commons Atribution 4.0 International.

Vol. 12/ Núm. 3 2025 pág. 2508
INTRODUCTION
Dating apps is one of the longest chapters in the history of media forms of matchmaking
for both erotic and affective relationships. This search began with postal mail in the 19th century,
then with advertisements in the print media and the emergence of matchmaking agencies in the
1960s, some of which operated internationally. In the 1990s, television dating programs became
popular, such as the paradigmatic case in Argentina of “Yo me quiero casar, y usted?”, hosted by
Roberto Galán. In addition, in that same decade the first dating websites emerged, with a
remarkable impact, as evidenced by the fact that in 1998 5% of heterosexual cisgender couples in
the United Kingdom had met online (Abad García et al., 2011; Allport, 1954).
On the other hand, information and communication technology (ICT)-mediated
sociosexual experiences have become part of everyday life for both men and women who have
same-gender sex in urban contexts. Both men and women value the sexual freedom and autonomy
that dating apps provide, while ensuring their physical and online safety (Parra & Obando, 2019).
These platforms not only facilitate contact with potential sexual partners, but also promote social
interaction between people with similar sexual preferences, in a specific and relatively private
environment that allows escaping heteronormative surveillance through various concealment or
anonymity strategies (Blackwell et al., 2015).
For more than a decade, studies on online encounters between gay men have been
conducted in English-speaking countries, mainly from behavioral and public health perspectives
(Chetcuti-Osorovitz, 2016). In Latin America, research on sexuality in digital environments has
gained popularity with the increase in Internet use and its influence on sociocultural processes.
Beginning in the late 2000s, texts in Spanish address topics such as sexual dissidence, subversive
gender manifestations, trans epistemologies, and virtual cruising, among others (Fonseca &
Quinter, 2009; Limón Piris, 2017).
In addition, location-centric social networks for dating, known as Mobile Mediated
Relationships (MMR), have experienced exponential growth in recent years. While these
platforms have a variety of uses, much of their popularity is due to the ability to have casual
encounters. Many of these face-to-face encounters, known as face-to-face (FTF), occur after
initially meeting through an MMR. Although there have always been means of real-life contact
(IRL) such as advertisements or phone lines, new online dating apps have radically transformed
the way people look for a partner or sex. To analyze this change, focused on the gay community,
using two of the most popular mobile dating apps among this group are highlighted, such as
Grindr and Tinder (Ato et al., 2013).
It is worth noting, that in the case of Grindr©, it is an application designed to facilitate
casual encounters, also known as hook ups. It could best be described as a “sociosexual”
networking platform, as it highlights interpersonal communication processes between those who

Vol. 12/ Núm. 3 2025 pág. 2509
are open to establishing connections of an erotic, platonic and practical nature, sometimes
simultaneously. The emergence and popularity of Grindr© respond to the needs of a generally
discriminated and stigmatized sector of the population, who find in this application an opportunity
to socialize with other men without exposing themselves to the dangers of the everyday world
(Sanz Menéndez, 2003).
In addition, we have Tinder, whose name derives from “dry and flammable material, such
as wood or paper, used to light a fire”. This definition is reflected in the company's logo, a flame,
which evokes the idea of a romantic spark between the two people matched, known as a “match,”
which is more associated with heterosexual relationships ( Aunspach, 2015). According to theorist
VanderMolen, heterosexual people often prefer a traditional approach to dating, where traditional
gender roles are maintained, such as the man assuming the role of initiator of verbal contact and
the woman following his lead. VanderMolen's observation, as well as Aunspach's assumption of
heterosexual-only romantic relationships, could be related to the name of the app (Viladrich et
al., 2017).
Until now, apart from the difference in sexual preferences, another significant distinction
between these two applications was that Tinder required users to select their photos and
information from their Facebook account, which had a significant impact on their level of trust.
Today, however, Tinder, like Grindr, has eliminated the requirement to link the account with
Facebook. In addition, the process of selecting potential partners differs between the two
platforms. While on Grindr you can send messages to all the profiles that appear, on Tinder a
mutual acceptance of both parties is required to start a conversation. Grindr displays twelve
profiles on the first screen, while on Tinder each profile is displayed individually. In terms of
relationships, Grindr tends to be more immediate than Tinder. In terms of images, Tinder allows
viewing in larger size and uploading up to five images, while Grindr only allows a single image
(Antheunis et al., 2007).
In the same way that new gender and body identities have emerged that do not fit the
traditional categories in networking sites, certain phobias have also emerged that represent a threat
to many of the freedoms won by the LGBT community. This refers specifically to transphobia
and plumophobia. Faced with these facts, the company responsible for Grindr has responded
through the series of episodes #kindr on its Youtube channel, where it gives voice to users who
feel discriminated against in the application due to sexual racism, transphobia or for not meeting
current beauty standards. In this same context, the non-profit campaign #stopplumophobia has
emerged in Spain, which seeks to make visible and raise awareness about the damage caused by
“lgtbiphobia” within the LGBT community itself (Attrill, 2012).
Therefore, in several countries around the world, the LGBT community (lesbian, gay,
bisexual, transgender and other sexual orientations different from heterosexuality) faces
discrimination, prejudice and social marginalization. Despite the historical advances made by

Vol. 12/ Núm. 3 2025 pág. 2510
members of this community in favor of equality, society still tends to consider heterosexuality
and conformity to the gender identity assigned at birth as social and cultural norms. This
phenomenon, known as “heteronormativity,” contributes to stigmatization and prejudice towards
LGBT people in society, generating a negative valuation based on their sexual orientation
(Francia Martínez et al., 2017; Fernández Rodríguez & Squiabro , 2014).
The study of online dating has grown in academic research on LGBT and discrimination,
exploring topics such as self-representation, communication, and authenticity in mobile apps
(Dawson & McIntosh, 2006; Antheunis et al., 2007; Gibbs et al., 2011). Representations of
identity, gender and sexual differences in online dating have been analyzed (Morgan et al., 2010).
In Spain, studies on social network analysis, new trends in networks and usefulness in
collaborative learning stand out (Flores Vivar, 2009). Gender representation in location-based
dating applications has also been investigated (Bostwick et al., 2014).
Scientific research confirms that discrimination, directed toward both the LGBT
community and other marginalized groups, is a pervasive social phenomenon that has serious
repercussions on the mental health of those affected (Burgess et al., 2007) (Woodford et al., 2015).
Although the detrimental impact on psychological well-being is recognized, in Mexico there is
no specific scale to assess perceived discrimination by LGBT individuals, especially as it relates
to the use of dating apps. The creation and validation of such a scale would be very useful for the
scientific community, not only in Mexico but worldwide, as it would provide information about
the experiences of this group and facilitate the implementation of organizations, institutional
programs and public policies that support the LGBT community (Bostwick et al., 2014).
Therefore, with this research we sought to develop and validate an instrument to measure
perceived discrimination at the individual and group level within the LGBTQ collective within
dating apps. It is intended to verify whether the psychometric properties of the instrument support
its usefulness in both clinical and research contexts. In addition, we seek to determine whether
the factorial structure of the instrument reflects the two proposed dimensions: individual
discrimination and group discrimination. To achieve this, analyses of the psychometric properties
of the instrument were carried out, with the aim of providing the scientific community with a
valid and reliable tool that contributes to the study of this phenomenon in a Mexican city, the
same country and Latin America.
METHODOLOGY
Design
This study adopts an instrumental research design Ato et al., (2013), focused on analyzing
the psychometric properties of the Discrimination Scale in dating apps within the LGBT
community, with the purpose of validating it in the city of Chiapas, Mexico.

Vol. 12/ Núm. 3 2025 pág. 2511
Participants
A non-probabilistic sample composed of 640 individuals belonging to the LGBT
community in the city of Chiapas, Mexico was used. The inclusion criteria established for
participation in the study were: (1) being at least 21 years of age, (2) residing in the city of
Chiapas, Mexico, and (3) identifying as part of the LGBT community.
RESULTS AND DISCUSSION
Instruments
A questionnaire was used to collect general data and to sociodemographically characterize
the participants. We inquired about their age, sexual orientation and marital status (Table 1).
Table 1
Sociodemographic information of the sample
Variable Category f %
Age
18-27 27 4,22
28-37 159 24,84
38-47 296 46,25
48-57 148 23,23
58 or more 10 1,56
Relationship
status
Single 233 36,61
Engagement 56 8,75
Married 22 3,44
Cohabitation 329 51,41
Sexual
orientation
Gay 89 13,91
Lesbian 36 5,63
Bisexual 313 48,91
Transgender 146 22,81
Other 56 8,75
Note. N:640
The LGBT Dating Discrimination Scale was designed by researchers to assess
discrimination experienced by individuals who identify as part of the LGBT community. The
instrument consists of 30 statements are distributed in seven dimensions: General context (7
items, not included in the validation) Discrimination outside the app (4 items, 3 included in the
validation), Interactions within the app (5 items, 4 included in the validation), Sexual racism in
the app (3 items, all included in the validation), Discrimination based on physical appearance (3
items, all included in the validation), Motivations and actions within the application (4 items, 3

Vol. 12/ Núm. 3 2025 pág. 2512
included in the validation), and Results of the use of the application (3 items, all included in the
validation). The 17 items subject to validation use a five-point Likert scale, ranging from 1
(Never) to 5 (Every day).
Procedures
Data collection was conducted through questionnaires on the Google Forms platform, using
dating apps (Grindr, Tinder, Wapa, Manhunt, and Every) to recruit participants. An informed
consent form was implemented that included (a) the purpose of the study, (b) the voluntary nature
of participation, (c) potential risks and benefits, (d) the participant's right to withdraw, (e)
institutional affiliation, and (f) the researchers' contact details. Subsequently, the collected data
were analyzed using R version 4.3.3, performing descriptive analysis, exploratory factor analysis,
item discrimination analysis, reliability analysis, comparative analysis and correlation analysis
between factors.
For the exploratory factor analysis, the principal axis extraction method with orthogonal
rotation was used to identify the latent variables underlying the items. This approach was selected
for two main reasons: first, the principal axis extraction method does not require assumptions of
normality Fabrigar et al., (1999), and second, orthogonal rotation is suitable for this purpose. To
determine the number of factors, two criteria were used: the sedimentation plot (Scree Test) and
the amount of variance explained by the extracted factors. The sedimentation plot seeks to identify
the factors whose associated values are large enough to be considered. The inflection point on the
graph indicates the number of factors to be extracted. Regarding the second criterion, Hatcher
(1994) suggests considering factors that explain at least 5% of the variance. In this study, two
factors met these criteria. As an acceptance criterion, items with a factor loading greater than .50
on a single factor were considered.
To evaluate the discriminatory capacity of the factor items, the item-total correlation index
(rbis) was calculated. Those items whose values were within the range of .30 to .70 were
considered appropriate (Field, 2013). To determine the reliability of the factors, Cronbach's alpha
coefficient and the Spearman-Brown split-half were calculated. DeVellis & Thorpe, (2017) states
that indices above .70 are acceptable, while those between .80 and .90 are considered good.
However, he suggests that if the alphas are greater than .90, consideration should be given to
revising the scale and reducing its length. On the other hand, Campo-Arias & Oviedo, (2008),
point out that indices above .90 may indicate redundancy or duplication of items, which implies
that at least a couple of items measure the same aspect of the construct and one of them should
be eliminated.
To carry out the confirmatory factor analysis, the R statistical program was used. The
maximum likelihood estimation method was employed together with the corrections of Satorra
and Bentler (Fingerhut et al., 2010). To evaluate the model fit, several goodness-of-fit indices
were used, such as the chi-square (χ2), the root mean squared error of approximation (RMSEA),

Vol. 12/ Núm. 3 2025 pág. 2513
the Tucker-Lewis index (TLI) and the comparative fit index (CFI). For the model to be considered
acceptable, CFI and TLI values were required to be ≥ .90, and RMSEA values were required to
be ≤ .08 (Byrne, 2010) (Hu & Bentler, 1999). In addition, it was established that the regression
coefficients of each item in its respective factor should exceed .50 (Hair, 2006).
Validation of the discrimination scale in dating apps in the LGBT community
Exploratory data analysis
In order to be able to understand the relationships between the values obtained during the
application of the instrument, an exploratory data analysis was performed, for which a frequency
distribution of all items was constructed, as well as the main statistical measures were calculated.
Table 2 contains the frequency distribution for the 17 items of the instrument that will be
analyzed with the psychometric tools. It is observed that, for items 14 to 16, the predominant
category is Almost every day, with absolute frequencies ranging from 282 (item 15 - 44.06% of
the sample) to 299 (item 16 - 46.72% of the sample) cases; while, for the other items, the modal
class is Every day, with frequencies ranging from 291 cases (item 23 - 45.47% of the sample) to
386 cases (item 27 - 60.31% of the sample).
Table 2
Frequency distribution by item
1 2 3 4 5
fa % fa % fa % fa % Fa %
It14 11 1.72 73 11.41 130 20.31 285 44.53 141 22.03
It15 15 2.34 37 5.78 121 18.91 282 44.06 185 28.91
It16 11 1.72 55 8.59 125 19.53 299 46.72 150 23.44
It17 5 0.78 29 4.53 86 13.44 179 27.97 341 53.28
It18 5 0.78 30 4.69 94 14.69 171 26.72 340 53.13
It19 10 1.56 17 2.66 54 8.44 179 27.97 380 59.38
It20 2 0.31 3 0.47 71 11.09 215 33.59 349 54.53
It21 6 0.94 49 7.66 100 15.63 157 24.53 328 51.25
It22 3 0.47 19 2.97 84 13.13 205 32.03 329 51.41
It23 7 1.09 39 6.09 117 18.28 186 29.06 291 45.47
It24 8 1.25 14 2.19 75 11.72 169 26.41 374 58.44
It25 4 0.63 30 4.69 85 13.28 179 27.97 342 53.44
It26 7 1.09 27 4.22 95 14.84 173 27.03 338 52.81
It27 8 1.25 19 2.97 55 8.59 172 26.88 386 60.31

Vol. 12/ Núm. 3 2025 pág. 2514
It28 3 0.47 4 0.63 72 11.25 212 33.13 349 54.53
It29 7 1.09 49 7.66 100 15.63 160 25.00 324 50.63
It30 4 0.63 20 3.13 84 13.13 205 32.03 327 51.09
Table 3 shows the summary statistics of the items analyzed. It is reported that the highest
average corresponds to item 27, with a value of 4.42, while the lowest average is item 14 with a
value of 3.74. Analyzing the dispersion of the data (measured through the standard deviation), it
is reported that this value ranges from 0.73 (item 20) to 1.02 (item 29). All the reported skewness
coefficients (measuring the asymmetry of the data) are negative, indicating that the general
tendency of the data is to accumulate on the right side of the spectrum, that is, on the higher values
of the scale; thus, all the reported averages have values higher than the central value of the scale.
To determine the shape of the curve of the data, the kurtosis was checked; only item 14
reports a negative kurtosis, which means that the data present a lower proportion of outliers
compared to a normal distribution; while the rest of the items have positive values, which means
that these items present a sharper frequency distribution (i.e., with a greater tendency to report
more extreme outliers than a normal distribution). All the items analyzed report a positive item-
total correlation, which indicates that all the items that make up the data set analyzed grow or
decrease in the same direction, the range of values reported is within the acceptable range (Ato et
al., 2013; Campo-Arias & Oviedo, 2008).
Table 3
Summary statistics by item
Media Desv Est Sesgo Curtosis CIT
It14 3.74 0.98 -0.61 -0.17 0.69
It15 3.91 0.96 -0.87 0.61 0.64
It16 3.82 0.95 -0.73 0.22 0.66
It17 4.28 0.92 -1.19 0.79 0.67
It18 4.27 0.93 -1.14 0.58 0.68
It19 4.41 0.87 -1.71 2.99 0.63
It20 4.42 0.73 -1.09 0.94 0.65
It21 4.18 1.02 -1.01 0.03 0.71
It22 4.31 0.84 -1.12 0.80 0.67
It23 4.12 0.98 -0.90 0.03 0.70
It24 4.39 0.87 -1.50 2.10 0.64
It25 4.29 0.91 -1.18 0.70 0.68
It26 4.26 0.94 -1.18 0.78 0.68
It27 4.42 0.86 -1.67 2.74 0.64

Vol. 12/ Núm. 3 2025 pág. 2515
It28 4.41 0.75 -1.18 1.35 0.63
It29 4.16 1.02 -1.01 0.07 0.71
It30 4.30 0.86 -1.15 0.93 0.64
Reliability Analysis
To verify the reliability of the model, the Cronbach's Alpha, the two halves and omega
models were used, the results are shown in Table 4 Verifying the Cronbach's Alpha values, it is
observed that the standardized coefficient is .921, the 95% confidence interval is within the range
considered acceptable. Through the method of the two halves, values between .846 and .964 are
reported with an average of 0.918, being similar to those obtained with Cronbach's Alpha.
Analyzing the value of the omega coefficient, it is reported that the model is able to explain
96.70% of the variability, the analysis of the hierarchical omega reveals that 84.4% of the
variability is attributable to the general factor of each model; all the values mentioned are within
the range considered acceptable as stated by both (Viladrich et al., 2017) and (Campo-Arias &
Oviedo, 2008).
Table 4
Summary of reliability measures
Method Indicator Value
Alfa
Cronbach
non-standardized .921
standarized (λ3) .921
IC 95%
lower .912
Media .921
Superior .930
two halves
Maximum (λ4) .964
Media .918
Mínimum (β) .846
Quantiles
2.5% .881
50.0% .918
97.5% .955
Omega hierarchical (ωh) .844
Total (ωt) .967
Figure 1 represents the histogram of the 24310 reliability values obtained through the two
halves method. It can be seen that the group with the highest absolute frequency is the one between
0.915 and 0.920, with 3036 cases, representing 12.49% of the sample; followed by the group of

Vol. 12/ Núm. 3 2025 pág. 2516
0.930-0.935, with 2886 cases reported, equivalent to 11.87% of the values obtained by the
method.
In summary, all the methods used reported values higher than the minimum
recommended by the specialized literature (DeVellis & Thorpe, 2017). Which allows affirming
that the instrument has the properties of stability, reproducibility and consistency, so it can be
used in the measurement of perceived discrimination in dating apps by the LGBT community. On
the other hand, the values of the correlations of each item with respect to the total score report
good internal consistency, which indicates that the items of the instrument are able to distinguish
subjects with different levels of perceived discrimination (Herrero-Jiménez & Caballero-Gálvez,
2017; González-Rivera & Pabellón-Lebrón, 2018).
Figure 1
Frequency distribution of reliability - two halves method
To measure the adequacy of the data through sampling adequacy, a KMO test is developed,
the average value is 0.867, with values ranging from 0.804 (item 17) to 0.985 (item 23); based on
the figures obtained, it is concluded that the analyzed data set has a degree of common variance
that can be considered meritorious and therefore a principal component analysis (PCA) can be
performed (Keith, 2015).
Principal Component Analysis
Table 5 shows the result of the Principal Component Analysis for the analyzed variable; it
is reported that the first five components report factor loadings higher than 1, so it is decided to
use that amount of components to develop a model; however, given the value close to 1 of the
sixth component, a second model with that amount of components will be tested; for the first case
(5 components), the model is able to explain 71. 06% of the variability, while the 6-component
model manages to agglutinate 76.90% of the natural dispersion of the data; both values are
considered acceptable according to the literature consulted (Fabrigar et al., 1999; Wu et al., 2007).

Vol. 12/ Núm. 3 2025 pág. 2517
Table 5
Principal Component Analysis
Component
Principal Component Analysis
factorial
loading
explained
variance
cumulative
variance
PC01 7.54 44.34 44.34
PC02 1.26 7.39 51.72
PC03 1.19 6.99 58.72
PC04 1.08 6.37 65.08
PC05 1.02 5.98 71.06
PC06 0.99 5.84 76.90
PC07 0.79 4.67 81.57
PC08 0.62 3.66 85.24
PC09 0.55 3.21 88.45
PC10 0.53 3.09 91.54
PC11 0.48 2.81 94.35
PC12 0.22 1.28 95.63
PC13 0.20 1.15 96.78
PC14 0.18 1.04 97.82
PC15 0.16 0.93 98.75
PC16 0.12 0.68 99.42
PC17 0.10 0.58 100.00
An orthogonal rotation with the varimax method was used to minimize the number of
variables within each factor (Field, 2013; Byrne, 2010); the result is shown in Tables 6 (5-
component model) and 7 (6-component model). For the case of the 5-component model, the
model coefficients range from 0.262 (RC5) to 0.335 (RC1), while, for the 6-component model,
the range is from 0.235 (RC6) to 0.297 (RC1).
Table 6
Orthogonal Rotation Results - 5-Dimensional Model
Media Desv Est Minimum Maximum
RC1 0.335 0.232 0.154 0.796
RC2 0.303 0.239 0.062 0.873
RC3 0.295 0.239 0.121 0.885
RC4 0.288 0.240 0.078 0.877
RC5 0.262 0.242 0.103 0.892

Vol. 12/ Núm. 3 2025 pág. 2518
Table 7
Orthogonal Rotation Results - 6-Dimensional Model
Media Desv Est Minimum Maximum
RC1 0.297 0.238 0.134 0.886
RC2 0.292 0.239 0.123 0.881
RC3 0.281 0.241 0.116 0.892
RC4 0.259 0.242 0.099 0.892
RC5 0.261 0.240 0.110 0.884
RC6 0.235 0.241 0.111 0.875
Analysis of the models
Finally, to evaluate the quality of the developed models, some goodness-of-fit indexes are
calculated, whose values are reported in Table 8. It is reported that, for both models, all the
calculated parameters are within the ranges recommended by the specialized literature (Abad
García et al., 2011; Hu & Bentler, 1999; Li, 2016): considered as normal or regular, so it is stated
that the developed models have a consistent or valid structure; however, the 6-component model
reports better values than the 5-component model, which is why the first one is selected.
Table 8
Goodness-of-fit indices for the models developed
Indicator n=5 n=6
CFI 0.841 0.912
TLI 0.797 0.885
RMSEA
Media 0.127 0.097
IC90-low 0.120 0.091
IC90-upper 0.134 0.104
SRMR 0.126 0.125
Once the number and composition of the dimensions of the instrument were obtained, we
proceeded to verify the behavior patterns of each of the dimensions. Comparing the means, it is
reported that Dimension 02 has the highest average of the group, with a value of 4.48, while
Dimension 05 reports the lowest average, with a value of 4.26. All dimensions, as well as the total
score, report bias coefficient with negative values, which indicates that the general tendency is
the predominance of high values, that is, the data are positioned in such a way that the average is
higher than the median. Verifying the kurtosis values, it is reported that Dimension 04 reports a
negative value, which indicates that the general tendency of the data is to have a more flattened

Vol. 12/ Núm. 3 2025 pág. 2519
or flattened structure than the normal distribution, while the rest of the dimensions present a
distribution of data with a greater concentration of values around the mean (Table 9).
Table 9
Summary statistics by dimension
Dim01 Dim02 Dim03 Dim04 Dim05 Dim06 Total
Media 4.30 4.48 4.31 4.42 4.26 4.27 4.44
Desv Est 0.91 0.79 0.81 0.70 0.96 0.90 0.71
Sesgo -1.22 -1.78 -1.00 -0.81 -1.08 -1.06 -0.98
Curtosis 0.90 3.60 0.35 -0.46 0.25 0.35 0.03
CIT 0.75 0.70 0.63 0.62 0.74 0.68 ---
Segmentation analysis
In order to detect possible behavioral patterns, a segmentation analysis was developed,
determining that the optimal number of segments is 2, Table 10 shows the summary statistics. It
is observed that segment 1 reports a mean of 3.706 and a standard deviation of 0.489, while
segment 2 has a higher mean (4.952) and a lower standard deviation (0.213) compared to segment
1. Figure 2 shows the differences in data density for both segments, a test of difference of means
reveals the existence of a significant difference (t= -38.788, p<0.05) in the levels of discrimination
for both segments. For this reason, individuals belonging to segment 2 can be referred to as
“individuals with a high level of perceived discrimination” in the sense that they report very high
levels of discrimination, either towards themselves or towards other people they know. This
higher level of perception group is often characterized by higher levels of anxiety and stress, as
well as other psychophysical manifestations (Stacey & Forbes, 2022; Wu & Trottier, 2022).
Table 10
Frequency distribution by segment
Segment 1 Segment 2
1 0 0
2 4 0
3 69 0
4 189 18
5 0 360
Total 262 378
Media 3.706 4.952
Desv Est 0.489 0.213
Sesgo -1.301 -4.265
Curtosis 0.566 16.280
Vol. 12/ Núm. 3 2025 pág. 2520
Figure 2
Discrimination Level by Segment
To verify whether the age range is a predictor of the level of discrimination, an analysis of
variance was performed, the results of which are shown in Table 11 and Figure 3; the analyses
show that there are no significant differences (F=0.568, p>0.05) between age groups for
discrimination, i.e., the age group is not a determinant of the level of discrimination experienced
in social networks, which is consistent with that reported by works such as (Luiggi-Hernández et
al., 2015) or (Lauckner et al., 2019).
Table 11
Analysis of Variance - Age Discrimination
degrees
freedom
Sum
squares
square
Medium F p-valor
Age 4 1.1 .2853 .568 0.686
Residual 635 318.7 .5019
Figure 3
Analysis of Variance - Discrimination by Age Group

Vol. 12/ Núm. 3 2025 pág. 2521
In practical terms, it could be demonstrated that the scale of discrimination in dating
applications in the LGBT community (EDDAC-LGBT) can be used for the execution of new
works in this area; which constitutes an advance considering the scarce instruments developed for
the Latin American area. Although the phenomenon of LGBT discrimination and its
repercussions on the physical and mental health of individuals have been investigated since the
1950s with works such as Lewin, (1952) and Allport, (1954), the literature consulted does not
report a specific instrument to measure the level of discrimination, however, all the literature
consulted reports the impact that discrimination has on aspects such as psychological well-being,
personal dissatisfaction and self-rejection, altering levels of stress, anxiety and depression
(Bostwick et al., 2014; González-Rivera & Pabellón-Lebrón, 2018).
CONCLUSIONS
Around the world, the LGBT community has been a constant victim of prejudice and
rejection in the social, occupational, academic and clinical spheres; despite this, there is no
specific instrument for the measurement of discrimination towards the LGBT community. By
virtue of meeting this need, this article presents the scale of discrimination in dating apps in the
LGBT community (EDDAC-LGBT), which reports psychometric properties suitable for its use
in the study sample, since it has been shown to have both validity and reliability. Although it was
possible to demonstrate through inferential statistics that the EDDAC-LGBT model is invariant
among the age groups considered, it is recommended that an analysis be made relating the values
obtained with the different sociodemographic variables. It is recommended that further studies be
conducted in other sample groups in order to corroborate the findings of this research and
determine their applicability in other regions of Mexico.

Vol. 12/ Núm. 3 2025 pág. 2522
REFERENCES
Aunspach, C. (2015). From the Gay Bar to the Search Bar: Promiscuity, Identity, and Queer
Mobility on Grindr. Tesis de Posgrado. University of Nebraska-Lincoln.
https://bit.ly/2RHTyTj
Abad García, F. J., Olea Díaz, J., Ponsoda Gil, V., & García García, C. (2011). Measurement in
Social and Health Sciences (Primera ed.). Madrid, España: Synthesis.
Allport, G. W. (1954). The nature of prejudice. Cambridge: Addison-Wesley.
Antheunis, M. L., Valkenburg, P. M., & Peter, J. (2007). Computer-mediated communication and
interpersonal attraction: an experimental test of two explanatory hypotheses.
Cyberpsychol Behav, 10(6), 831-835.
https://doi.org/https://doi.org/10.1089/cpb.2007.9945
Ato, M., López-García, J. J., & Benavente, A. (2013). A Classification System of Research
Designs in Psychology. Annals of Psychology, 29(3), 1038-1059.
https://doi.org/https://doi.org/10.6018/analesps.29.3.178511
Attrill, A. (2012). Sharing Only Parts of Me: Selective Categorical Self-Disclosure Across
Internet Arenas. International Journal of Internet Science, 7(1), 55-77.
https://www.researchgate.net/publication/267800740_Sharing_Only_Parts_of_Me_Sele
ctive_Categorical_Self-Disclosure_Across_Internet_Arenas
Blackwell, C., Birnholtz, J., & Abbott, C. (2015). Seeing and being seen: Co-situation and
impression formation using Grindr, a location-aware gay dating app. New Media Society,
17(7), 1117-1136. https://doi.org/https://doi.org/10.1177/1461444814521595
Bostwick, W. B., Boyd, C. J., Hughes,, T. L., West, B. T., & McCabe, S. E. (2014).
Discrimination and mental health among lesbian, gay, and bisexual adults in the United
States. American Journal of Orthopsychiatry, 84(1), 35-45.
https://doi.org/https://doi.org/10.1037/h0098851
Burgess, D., Lee, R., Tran, A., & Van Ryn, M. (2007). Effects of perceived discrimination on
mental health and mental health services utilization among gay, lesbian, bisexual and
transgender persons. Journal of LGBT Health Research, 3(4), 1-14.
https://doi.org/https://doi.org/10.1080/15574090802226626
Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications,
and programming (2a ed.). New York: Taylor & Francis Group.
https://doi.org/10.4324/9780203726532
Campo-Arias, A., & Oviedo, H. C. (2008). Psychometric Properties of a Scale: Internal
Consistency. Public Health Journal, 10(5), 831-839.
https://doi.org/https://doi.org/10.1590/s0124-00642008000500015

Vol. 12/ Núm. 3 2025 pág. 2523
Chetcuti-Osorovitz, N. (2016). Sexualities Among Women and Digital Uses. Historical
Sociology(6), 153-178. https://revistas.um.es/sh/article/view/278691
Dawson, B. L., & McIntosh, W. D. (2006). Sexual strategies theory and Internet personal
advertisements. Cyberpsychol Behav, 9(5), 614-617.
https://doi.org/https://doi.org/10.1089/cpb.2006.9.614
DeVellis, R. F., & Thorpe, C. T. (2017). Scale development: Theory and applications (Cuarta
ed.). California: Sage Publications.
Fabrigar, L. R., Wegener,, D. T., & Strahan, E. J. (1999). Evalua the use of exploratory factor
analysis in psychological research. Psychological Methods, 4(3), 272-299.
https://doi.org/https://doi.org/10.1037/1082-989x.4.3.272
Fernández Rodríguez, M. d., & Squiabro , J. C. (2014). Prejudice and Social Distance Toward
Homosexual Individuals Among University Students. Puerto Rican Journal of
Psychology, 25(1), 52-60.
http://pepsic.bvsalud.org/scielo.php?script=sci_arttext&pid=S1946-
20262014000100004&lng=pt&tlng=es
Field, A. (2013). Discovering statistics using IBM SPSS statistics . London, England.: SAGE.
Fingerhut, A. W., Peplau, L. A., & Gable, S. L. (2010). Identity, minority stress and psychological
well-being among gay men and lesbians. Psychology & Sexuality, 1(2), 101-114.
https://doi.org/https://doi.org/10.1080/19419899.2010.484592
Flores Vivar, J. M. (2009). New Communication Models, Profiles, and Trends on Social
Networks. To Communicate(33), 73-81. https://doi.org/https://doi.org/ 10.3916/c33-
2009-02-007
Fonseca Hernández, C., & Quintero Soto, M. L. (2009). Queer Theory: The Deconstruction of
Peripheral Sexualities. Sociological (Mexico), 24(69), 43-60.
Francia Martínez, M., Reyes, C. E., & Lespier, Z. (2017). Attitudes, Knowledge, and Social
Distance of Psychotherapists Toward the Transgender and Transsexual Community.
Puerto Rican Journal of Psychology, 28(1), 98-113.
https://www.redalyc.org/pdf/2332/233255999007.pdf
Gibbs, J. L., Ellison , N. B., & Lai, C.-H. (2011). First Comes Love, Then Comes Google: An
Investigation of Uncertainty Reduction Strategies and Self-Disclosure in Online Dating.
Communication Research, 38(1), 70-100.
https://doi.org/https://doi.org/10.1177/0093650210377091
González-Rivera, J. A., & Pabellón-Lebrón, S. (2018). Development and Validation of an
Instrument to Measure Perceived Discrimination in the LGBT Community. Evaluar
Journal, 18(2), 59-74. https://doi.org/https://doi.org/10.35670/1667-4545.v18.n2.20809
Hair, J. F. (2006). Multivariate Data Analysis (sexta ed.). New Jersey: Pearson international
edition.

Vol. 12/ Núm. 3 2025 pág. 2524
Herrero-Jiménez, B., & Caballero-Gálvez, A. A. (2017). Gender Representations in Mobile
Networking Apps. Social Prism(2), 31-56.
https://doi.org/https://revistaprismasocial.es/article/view/1549
Hu, L.‐t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6(1), 1-55. https://doi.org/https://doi.org/
10.1080/10705519909540118
Keith, T. Z. (2015). Multiple Regression and Beyond: An Introduction to Multiple Regression and
Structral Equation Modeling. New York: Routledge.
https://doi.org/https://doi.org/10.4324/9781315162348
Lauckner, C., Truszczynski, N., Lambert, D., Kottamasu, V., Meherally, S., Schipani-
McLaughlin, A. M., . . . Hansen, N. (2019). “Catfishing,” cyberbullying, and coercion:
An exploration of the risks associated with dating app use among rural sexual minority
males. Journal of Gay & Lesbian Mental Health, 23(3), 289-306.
https://doi.org/https://doi.org/ 10.1080/19359705.2019.1587729
Lewin, K. (1952). Field theory in social science. New York: Haper and Row.
Li, C.-H. (2016). Confirmatory factor analysis with ordinal data: Comparing robust maximum
likelihood and diagonally weighted least squares. Behavior Research Methods, 48, 936-
949. https://doi.org/https://doi.org/10.3758/s13428-015-0619-7
Limón Piris, B. (2017). From the Handkerchief to Grindr: An Artistic Analysis of Sexual Semiotic
Codes Among Men Who Have Sex with Men. Polytechnic University of Valencia.
http://hdl.handle.net/10251/97214
Luiggi-Hernández, J. G., Laborde Torres, G. E., González Domínguez, J., Carrasquillo Sánchez,
G. M., Piñero Meléndez, M., Castro Medina, D. M., & González Rentas, C. D. (2015).
Outing the discrimination towards LGBT people during the hiring process: What about
their well-being? Revista Puertorriqueña de Psicología, 26(2), 194-213.
https://www.researchgate.net/publication/283721881_Outing_the_Discrimination_Tow
ards_LGBT_People_During_the_Hiring_Process_What_About_Their_Well-Being
Morgan, E. M., Richards, T. C., & VanNess, E. M. (2010). Comparing narratives of personal and
preferred partner characteristics in online dating advertisements. Computers in Human
Behavior, 26(5), 883-888. https://doi.org/https://doi.org/10.1016/j.chb.2010.02.002
Parra, L., & Obando, A. (2019). From Virtual Closets to Digital Extermination Fields:
Intersectionality of Grindr and the Fabrication of Chilean Gay Identity. Communication
and Media, 28(40), 98-113. https://doi.org/https://doi.org/10.5354/0719-
1529.2019.54008

Vol. 12/ Núm. 3 2025 pág. 2525
Sandfort TG, D. G. (2003). Same-sex sexuality and quality of life: Findings from the Netherlands
Mental Health Survey and incidence study. Archives of Sexual Behavior, 15-22.
https://doi.org/http://dx.doi.org/10.1023/A:1021885127560
Sanz Menéndez, L. (2003). Social Network Analysis: Or How to Represent Underlying Social
Structures. Notes on Science and Technology(7), 21-29. http://hdl.handle.net/10261/1569
Stacey, L., & Forbes, T. D. (2022). Feeling Like a Fetish: Racialized Feelings, Fetishization, and
the Contours of Sexual Racism on Gay Dating Apps. The Journal of Sex Research, 59(3),
372-384. https://doi.org/http://dx.doi.org/10.1080/00224499.2021.1979455
Viladrich, C., Angulo-Brunet, A., & Doval, E. (2017). A journey around alpha and omega to
estimate internal consistency reliability. Annals of Psychology, 33(3), 755-782.
https://doi.org/http://dx.doi.org/10.6018/analesps.33.3.268401
Woodford, M. R., Paceley, M. S., Kulick, A., & Jun Sung, H. (2015). The LGBT social climate
matters: Policies, protests, and placards and psychological well-being among LGBT
emerging adults. Journal of Gay & Lesbian Social Services, 27(1), 116-141.
https://doi.org/https://doi. org/10.1080/10538720.2015.990334
Wu, A. D., Li, Z., & Zumbo, B. D. (2007). Decoding the meaning of factorial invariance and
updating the practice of multigroup confirmatory factor analysis: A demonstration with
TIMSS data. Practical Assessment Research & Evaluation, 12(3), 1-26.
https://doi.org/https://doi.org/10.7275/mhqa-cd89
Wu, S., & Trottier, D. (2022). Dating apps: A literature review. Annals of the International
Communication Association, 46(2), 91-115.
https://doi.org/https://doi.org/10.1080/23808985.2022.2069046