ISSN 0214 - 9915 CODEN PSOTEG

Psicothema 2020, Vol. 32, No. 2, 253-260

Copyright © 2020 Psicothema
doi: 10.7334/psicothema2019.344 www.psicothema.com

Frequency of Suicidal Ideation Inventory: Psychometric Properties

of the Spanish Version
Nicolás Sánchez-Álvarez1, Natalio Extremera Pacheco1, Lourdes Rey1, Edward C. Chang2, and Olivia D. Chang2
1
University of Málaga (Spain) and 2 University of Michigan, USA

Abstract Resumen
Background: The Frequency of Suicidal Ideation Inventory (FSII) is a Inventario de Frecuencia de Ideación Suicida: Propiedades Psicométricas
new and brief measure designed to assess for suicide ideation frequency de la Versión Española. Antecedentes: el Inventario de Frecuencia
in the previous year. To provide evidence of the reliability and validity for de Ideación Suicida (FSII) es una medida nueva y breve que evalúa la
the FSII in a Spanish-speaking community, the present study examined frecuencia de ideaciones suicidas en el último año. Para proporcionar
the psychometric properties of the Spanish version of the FSII (FSII-S) evidencias de fiabilidad y validez del FSII en la comunidad hispano-
in a sample of 1,013 adults from southern Spain (51.4% women), ranging hablante, el presente estudio examinó las propiedades psicométricas de
from 18 to 82 years old. Method: In addition to the FSII-S, the participants la versión española del FSII (FSII-S) en una muestra de 1.013 adultos del
completed the following measures: Suicidal Behaviors Questionnaire- sur de España (51,4% mujeres), con un rango de edad de 18 a 82 años.
Revised, Beck Depression Inventory, Positive and Negative Affectivity Método: además del FSII-S, los participantes completaron las siguientes
Scale, Flourishing Scale, and Satisfaction with Life Scale. Results: The medidas: Cuestionario Revisado de Comportamientos Suicidas, Inventario
results of Confirmatory Factor Analysis indicated a good fit for a one- de Depresión de Beck, Escala de Afectividad Positiva y Negativa, Escala
dimensional model of the FSII-S, and high reliability (a = .89). FSII-S de Plenitud y Escala de Satisfacción Vital. Resultados: los resultados de
scores correlated negatively with suicide protective factors, and positively la estructura interna indicaron un buen ajuste del modelo unidimensional
with suicide risk factors. Multi-group analyses showed the invariance of del FSII-S y una alta fiabilidad (a = .89). Las puntuaciones del FSII-S
the factor structure of FSII-S across gender and age. Conclusions: In correlacionaron negativamente con factores protectores del suicidio y
line with studies in other countries, these findings provide evidence of positivamente con factores del riesgo suicida. Los análisis multigrupos
the validity, reliability, and invariance of the FSII-S as a brief measure of mostraron invarianza de estructura factorial del FSII-S por género y
suicidal ideation frequency in Spanish adults. edad. Conclusiones: consistente con estudios en otros países, estos
Keywords: Suicidal ideation, Spanish validation, measurement, factorial hallazgos proporcionan evidencia de la validez, fiabilidad e invarianza del
validity, adult population. FSII-S como medida breve de frecuencia de ideación suicida en adultos
españoles.
Palabras clave: ideación suicida, validación española, medición, validez
factorial, población adulta.

According to official figures reported by the World Health recent studies have found that the prevalence of deaths by suicide
Organization, in 2017, nearly 800,000 people died by suicide in Spain in 2017 was 87 per 100,000 individuals (World Health
worldwide. Suicide accounted for 1.4% of all deaths globally, Organization, 2018), with some research reporting a relative
making it one of the leading causes of death among individuals increase in suicides, especially in the working-age adult population
(World Health Organization, 2018). This figure is even greater if (López-Bernal, Gasparrini, Artundo, & McKee, 2013). In Spain,
one includes the number of suicide attempts made. Notably, after remarkable suicide rate trends from 1974/1976 to 1999/2000 have
the 2008 economic crisis, rates of suicide increased in the US been reported, in which rates increased considerably (Hansen &
and European countries (Chang, Stuckler, Yip, & Gunnell, 2013), Pritchard, 2008). In short, several studies have shown that adults
and there were increases in negative mental health consequences aged 35 to 54 years are the most affected in both sexes, especially
in Andalusia, the most populated region of Spain (Córdoba-Doña,
in Spain (Gili, Roca, Basu, McKee, & Stuckler, 2013). Indeed,
Sebastián, Escolar-Pujolar, Martínez-Faure, & Gustafsson, 2014).
It is important to identify people who are at risk of dying by suicide
Received: November 6, 2019 • Accepted: February 10, 2020 by developing useful tools that can better identify risk of suicide
Corresponding author: Nicolás Sánchez-Álvarez to prevent suicide among vulnerable Spanish adult non-clinical
Faculty of Psychology populations (García-Nieto et al., 2012).
University of Málaga (Spain)
29071 Málaga (Spain) The number of previous suicide attempts is considered one
e-mail: nsa@uma.es of the most important predictors of a subsequent suicide attempt

253
 Nicolás Sánchez-Álvarez, Natalio Extremera Pacheco, Lourdes Rey, Edward C. Chang, and Olivia D. Chang

(Fernández-Artamendi et al., 2019; Goñi-Sarriés, Blanco, Azcárate, factor. Proximal risk factors are more closely linked to suicidal
Peinado, & López-Goñi, 2018). Negative life events, low self- behavior and often act as precipitants (e.g., suicide ideation).
esteem and stressful circumstances are known to trigger suicide risk, However, there are also some other positive aspects related
with suicide ideation being an important risk predictor for suicidal to well-being and personal traits which are seen as protective
behaviors and suicide attempts (Franklin et al., 2017; Soto-Sanz et factors, reducing the risk of suicide attempts (e.g., flourishing,
al., 2019), including among young Spanish adults (Blasco et al., life satisfaction, optimism, positive affect) (Hirsch, Duberstein,
2019). Suicidal ideation refers to thoughts of tiredness with one’s Chapman, & Lyness, 2007). Regarding sociodemographic
life, desires to take one’s own life as a solution to problems, and dimensions associated with suicide ideation measured by the FSII,
feelings that you have a life that is not worthwhile (Beck, Kovacs, as far as we know, no prior research has analysed the invariance by
& Weissman, 1979). These thoughts are often accompanied sex and gender groups. It is speculative to assume that differences
by an active desire to die and planning how to die, and these in prevalence of stressful events and specific life stressors across
thoughts can be recurring (Joiner, Rudd, & Rajab, 1997). Several gender and age stages might lead to differences between males and
instruments have been developed to assess and identify early signs females across the adult life span in psychiatric and mental health
of suicidal ideation in young adult populations (Batterham et al., outcomes (Hatch & Dohrenwend, 2007). Likewise, determining
2015; Blasco et al., 2019; García-Nieto, Blasco-Fontecilla, Paz- the invariance of factor structure across sex and gender is necessary
Yepes, & Baca-García, 2013); many of these instruments share to justify consistency of interpretation in suicide ideation for males
problems related to length and heterogeneity (Beck et al., 1979). and females in different age groups. Self-report instruments such
Current instruments which assess suicidal ideation have several as the FSII are assumed to measure a latent factor, which is used
methodological problems. For instance, some of the items on to compare groups or to follow individuals’ suicide ideation over
these scales evaluate other variables which are related to suicidal time. Therefore, the FSII should measure the identical construct
ideation, such as capability of attempt, courage, or competence with the same structure across different groups. One main aim of
(Beck et al., 1979), and even variables related to the method and this study was to analyze the measurement invariance of the FSII
ability to carry out suicidal behavior are evaluated (Beck, Brown, to demonstrate that individuals across gender and age interpret
& Steer, 1997). Therefore, some instruments assess latent factors the items and the underlying latent factors in the same way and,
other than suicidal ideation, which in turn influence a large number therefore, factor means can be compared across different groups.
of items (Batterham et al., 2015). In Spain, there exists validated, In summary, although there are well-validated Spanish scales
brief suicide scales, which have typically been used in Spanish to assess suicidal thoughts and attempts (Fonseca-Pedrero et
populations, such the Paykel scale (Fonseca-Pedrero et al., 2018; al., 2018; Rubio et al., 1998), there is a lack of brief and useful
Rubio, Montero, Salvador, & Marín, 1998). This scale is comprised measures that specifically assess the frequency of suicide ideation
of five items that assess jointly the presence of thoughts of death, in Spanish samples. It is important to validate new suicidal
suicide ideation, suicide plans, and suicide attempt in the last year. ideation measures in Spanish-speaking samples to identify early
Respondents answer each item “yes” or “no” which does not allow signs of suicide risk in adult populations. Likewise, providing
to quantify the frequency of suicidal ideation. measurement invariance across gender and age for the FSII has
Recently, Chang and Chang (2016) developed the Frequency practical significance for the use of this scale across different age
of Suicidal Ideation Inventory (FSII), which has been validated and gender groups and for the longitudinal tracking of participants.
in non-clinical populations of several countries (American, Thus, these instruments would help to examine the specific factors
Turkish, Hungarian, and Chinese samples) (Chang & Chang, associated with suicidality (e.g., suicidal ideation) in Spain, in order
2016). The goal of this scale is to assess the frequency of thoughts to create tailored suicide prevention programs among vulnerable
about suicide as an indicator of suicide risk in the general non- Spanish adult non-clinical populations and to potentially provide
clinical adult population. This measure is a five-item self-report information on developmental periods during which females and
questionnaire focused on one specific suicidal ideation dimension males are particularly vulnerable to suicide-related experiences
to evaluate the frequency of suicidal ideation over the last 12 (Hawton & van Heeringen, 2009).
months. Previous studies of the psychometric properties of the To our knowledge, a validated version of the FSII does not yet
FSII in different samples have confirmed a one-dimensional factor exist for use with Spanish adults. To fill this gap, the main aim of
structure (Chang & Chang, 2016; Chang et al., 2017). The English this study was to validate the Spanish version of the FSII scale in
version of the FSII has been well-validated, with good six-week a non-clinical sample of adults. Our main objectives were to: (a)
test–retest reliability. Different translations from China, Hungary, to examine the current frequency suicide ideation in a Spanish-
and Turkey have shown adequate internal reliability in several speaking adult sample. On the basis of prior findings in the original
international samples (Chang & Chang, 2016). English version and in other international non-clinical samples,
The FSII has been demonstrated to be positively correlated we expected a low average frequency of suicide ideation in our
with other measures of suicidal behaviors and risk factors, and Spanish sample; b) to analyse the factor structure and reliabilities
negatively correlated with suicide protective factors, which of the FSII-S. In line with prior studies, we expected that the factor
provides empirical evidence that supports factorial validity and structure analysis should confirm the one-dimensional structure,
construct validity (Chang & Chang, 2016). According to previous and an adequate internal reliability; (c) to examine the associations
studies, both distal factors (e.g., depressive symptoms, negative of the FSII-S with other factors related to suicidal ideation. We
affectivity, hopelessness) (Beck, Steer, Beck, & Newman, 1993) expected positive and significant associations between FSII-S scores
and proximal factors (e.g., suicidal ideation) are factors that and suicidal behaviors, depressive symptoms, negative affectivity,
may increase the risk of dying from suicide. Distal factors (e.g., and negative and significant associations with flourishing, life
depressive symptoms) affect the threshold for suicide and indirectly satisfaction, and positive affect; and (d) to examine factor structure
increase an individual’s risk when they experience a proximal risk and measurement invariance across gender and age. We expected

254
 Frequency of Suicidal Ideation Inventory: Psychometric Properties of the Spanish Version

that latent mean differences on the FSII-S are comparable across Positive and Negative Affect Schedule (PANAS; Watson, Clark,
gender and age groups. & Tellegen, 1988). This scale, developed by Watson et al. (1988)
was adapted for Spanish children and adolescents by Sandín
Method (2003). It is a self-report measure that includes 20 items of which
ten items refer to positive affect (PA; e.g., “I am inspired”) and
Participants ten items that refer to negative affect (NA; e.g., “I feel tense”).
Participants indicate how they feel and/or behave most of the time,
In this study, the sample was composed of 1,013 adults (521 using a scale ranging from 1 (never) to 3 (many times).
females) from southern Spain (Andalusia) with ages ranging from Flourishing Scale (FS; Diener et al., 2010). This scale is an eight-
18 to 82 years (M= 39.64, SD= 16.17). The sample was selected item measure that assesses several aspects of positive relationships,
from the University of Jaén (26.9% students) and the community, feelings of competence, and purpose in life. Participants indicate
using non-probability and convenience sampling. The total sample their agreement with each item on a seven-point Likert scale, from
reported marital status: single (42.3%), married (49.0%), divorced 1 (strongly disagree) to 7 (strongly agree). Higher scores indicate
(4.9%), and widowed (3.8%). Regarding the academic level, a positive self-view in important areas of functioning. We used
34.3% reported primary studies, 29.9% secondary studies, 29.1% the Spanish version, which has shown satisfactory psychometric
university studies, and 6.7% non-studies. Finally, the employment properties (Ramírez-Maestre et al., 2017).
status: employed (56.3%), unemployed (32.4%), and retired Satisfaction with Life Scale (SWLS; Diener, Emmons, Larsen,
(11.4%). The sample was split into three age groups: the first & Griffin, 1985). This scale is comprised of five self-referencing
group consisted of 329 young adults aged between 18 and 29 (M= statements about perceived global life satisfaction across a seven-
21.12, SD= 3.16, 60% female). The second group consisted of 361 point Likert scale. Participants indicate their extent of agreement/
middle-aged adults aged between 30 and 49 (M= 39.33, SD= 6.23, disagreement with each item on a scale of 1 (strongly disagree)
48% female). The third group consisted of 323 older adults aged to 5 (strongly agree). We used the well-validated Spanish version
between 50 and 82 (M= 58.84, SD= 6.97, 46% female). which has shown evidence of discriminant validity and adequate
internal consistency (Atienza, Pons, Balaguer, & Merita, 2000).
Instruments
Procedure
Sociodemographic data. All participants provided basic
sociodemographic data about their age, sex, academic level, Translation of the instrument was conducted according to
employment status, and marital status. recommended guidelines by Zenisky and Hambleton (2015):
Frequency of Suicidal Ideation Inventory (FSII; Chang & two authors of this study independently translated the original
Chang, 2016). The FSII is a 5-point scale that assesses the English FSII into Spanish. Then, a bilingual independent translator
frequency of suicidal thoughts over the past 12 months, from 1 performed the back translation. The discrepancies between the
(never) to 5 (almost every day). The total score ranges from 5 to 25, original version and the Spanish version were discussed and
with higher scores indicating greater suicidal ideation frequency adjusted by a panel of experts from the university.
(see Table 1). Data collection took place at a university. Undergraduate
Suicidal Behaviors Questionnaire-Revised (SBQ-R; Osman students enrolled in a class of methodology at the Faculty of
et al., 2001). The SBQ-R is a brief self-report instrument that Psychology. The principal researcher trained the students in the
evaluates four symptoms related to suicide. Adequate estimates fundamentals and instructions of collecting data using a snowball
of internal consistency and reliability have been reported in prior sampling strategy (Hendriks, Blanken, Adriaans, & Hartnoll,
work (Osman et al., 2001). The Spanish version has good levels of 1992). This strategy was adopted to distribute the battery of
reliability and validity in Spanish samples (Extremera, Quintana- questionnaires to other samples from the community, fostering
Orts, Mérida-López, & Rey, 2018). the heterogeneity of the sample in relation to gender, age, and
Beck Depression Inventory (BDI; Beck, Steer, & Garbin, 1988). educational levels (Demerouti & Rispens, 2014). Participants were
Used for the assessment of the intensity of depressive symptoms, informed of the confidentiality and anonymity of the survey, as
this scale consists of 21 items that assess negative mood, pessimism, well as the voluntary nature of participation. None of participants
weight loss, fatigue, suicidal thoughts, guilt, and sense of failure. received any reward for taking part in the study.
The total score ranges from 0 to 63. We used a well-validated Aware that a student-recruited sampling technique has several
Spanish version that has shown good internal consistency and limitations, we applied an administrative procedure to reduce
reliability (Sanz & Vázquez, 1998). the associated errors. First, inclusion and exclusion criteria

Table 1
Frequency of Suicide Ideation Inventory Spanish version (FSII-S)

Instructions: Please read carefully each statement about you and indicate the extent to which you agree or disagree [Instrucciones: Por favor lee los siguientes enunciados respecto a ti mismo e
indica tu grado de acuerdo o desacuerdo].
1. Over the past 12 months, how often have you thought about hurting yourself? [Durante los últimos 12 meses ¿con qué frecuencia has pensado en hacerte daño?]
2. Over the past 12 months, how often have you believed that your life was not worth living? [Durante los últimos 12 meses ¿con qué frecuencia has creído que no merecías vivir?]
3. Over the past 12 months, how often have you wondered what would happen if you ended your own life? [Durante los últimos 12 meses ¿con qué frecuencia te has preguntado qué pasaría si
pusieses fin a tu vida?]
4. Over the past 12 months, how often have you thought about committing suicide? [Durante los últimos 12 meses ¿con qué frecuencia has pensado en suicidarte?]
5. Over the past 12 months, how often have you wished you did not exist? [Durante los últimos 12 meses ¿con qué frecuencia has deseado no existir?]

255
 Nicolás Sánchez-Álvarez, Natalio Extremera Pacheco, Lourdes Rey, Edward C. Chang, and Olivia D. Chang

were applied: (1) participants over 18 years old, (2) not obtain with the previous one. For this, the chi-square difference test χ2,
informed written consent, and (3) Spanish people. Second, RMSEA, SRMR, and CFI are used. The variations in the fit of the
sociodemographic data were collected to get a stratification of sex, models are compared and are quantified through ΔCFI, ΔSRMR
age, and educational level. Secondly, the questionnaires included and ΔRMSEA. When sample size is high than 300, increases
explicit instructions to avoid the inconveniences of the application greater than ΔCFI < .010, ΔSRMR < .030 and ΔRMSEA < .015
of the snowball sampling technique. suggest the existence of invariance (Chen, 2007).

Data analysis Results

Statistical analyses were carried out using the Statistical Means, Standard Deviations, and Internal Consistencies
Package for the Social Sciences (SPSS; IBM, 2015) and Rstudio
(Rosseel, 2012). For Confirmatory Factor Analysis (CFA) and Means, standard deviations, and reliability of the FSII scores
Multiple Group Confirmatory Factor Analysis (MG-CFA) we from the present Spanish sample are shown in Table 2, together
used the Laavaan, semPlot, polycor, and semTools packages, with the scores obtained in US, Turkish, Hungarian, and Chinese
using the Weighted Least Squares Mean and Variance Adjusted samples of other studies (Chang & Chang, 2016). Spanish
method. Model fit was tested using χ2, the Comparative Fit Index participants reported low frequency of suicidal ideation (M= 7.87;
(CFI), the Tucker Lewis Index (TLI), Standardized Root Mean SD= 3.88). These findings are consistent with findings obtained
Square Residual (SRMR), and the Root Mean Square Error of in prior research in non-clinical adult populations (e.g., American
Approximation index (RMSEA), with values of CFI and TLI adults, Turkish adults, etc.). Internal consistency coefficients were
> .90, SRMR and RMSEA < .08, indicating a good fit (Kline, similar to those indicated in the original scale (α= .89).
2015). The percentage of missing data was acceptable (3.6%).
Thus, the missing values were imputed using the imputation Confirmatory Factor Analysis
algorithm of expectation–maximization (EM) (Liang & Bentler,
2004). To determine the internal consistency of the instruments, Given that prior research has found that the FSII taps into a
we estimated Cronbach’s alpha coefficient (α), McDonald’s single latent factor, we conducted a confirmatory factor analysis on
omega coefficient (ω), Guttman’s lambda (λ), and the function of the FSII-S. The index of fit of the factor structure model indicated
information from the Item Response Theory (Muñiz & Fonseca- a good fit, with the following indices: X2 (5, N= 1,013)= 1.102,
Pedrero, 2019). p= .954, TLI= .999, CFI= .999, SRMR= .017, and RMSEA= .001
To examine the factorial invariance across gender and age (CI 95%= .010-.050). Factor loadings (standardized) and item
groups, a MG-CFA was conducted through four levels of factorial information for the FSII-S are presented in Table 3.
invariance: configural invariance, metric factorial invariance, scalar
factorial invariance, and strict factorial invariance (Timmons,
2010). The invariance constraint test is based on the same Table 2
parameters which exist for all groups (configural invariance). If the Means, Standard Deviations, and Internal Consistencies of the Frequency of
Suicidal Ideation Inventory in Samples from Spain, US, Turkey, Hungary, and
invariance with no constraints model is confirmed, that means that China
equivalence exists across groups. Afterwards, the factor loadings
are required to be equivalent across groups (metric invariance). N M SD α
In the case where metric invariance is met, the different groups
Spanish sample 1013 7.87 3.88 .89
respond to items in the same way. Then, in the next step, the
US sample 399 6.35 3.26 .96
equivalence of means for items across groups (scalar invariance)
Turkish sample 153 7.87 3.65 .85
is included. Finally, the factor loadings, means of the indicators,
Hungarian sample 146 7.31 3.72 .93
residual variances, and residual covariances are constrained to be
Chinese sample 102 7.53 3.12 .85
equal across groups (strict invariance). The invariance of residual
variances and residual covariances showed equal latent scores Note: Means and Standard Deviations for US, Turkish, Hungarian, and Chinese samples as
across groups. This step invariance process compares each model reported in Chang and Chang (2016)

Table 3
Polychoric correlation, Factor loadings, Mean, SD, Skewness, Kurtosis, corrected Item-Total correlation for the FSII-S

Item 1 2 3 4 Factor loadings Mean(SD) Skewness Kurtosis r

Total Young Middle Older

Female Male
sample adults adults adults

1 – .827 .816 .839 .815 .877 .760 1.46(.83) 1.92 3.42 .83
2 .75 – .797 .843 .746 .779 .800 .799 1.60(.94) 1.58 1.98 .82
3 .71 .70 – .770 .771 .769 .761 .754 .825 1.80(1.05) 1.16 0.53 .83
4 .84 .76 .80 – .761 .767 .753 .737 .777 .752 1.39(.80) 2.36 5.52 .86
5 .75 .77 .74 .83 .856 .837 .877 .878 .825 .863 1.63(.96) 1.51 1.73 .86

N= 1013

256
 Frequency of Suicidal Ideation Inventory: Psychometric Properties of the Spanish Version

From the perspective of the Theory of Response to the Item, results’ invariance (Timmons, 2010) showed that differences
the complementary psychometric indicators were estimated using between gender or age in the model were non-significant (see
the WINMIRA program. The distribution and information function Table 5). The invariance tests by sex and age groups of the model
parameters of the FSII-S scale are presented in Figure 1. As can be begins with the establishment of the baseline model (model 1);
seen in the figure, the FSII-S test provides maximum information the model showed an adequate adjustment which supports the
for high values. configurational validity through sex and age groups. For metric
invariance, the results showed an appropriate adjustment of model
Convergent Validity 2, and the differences between models 2 and 1 did not exceed
the established threshold (ΔCFI < .010, ΔSRMR < .030, and
Pearson correlations for all variables are presented in Table
ΔRMSEA < .015), contributing to the metric invariance support
4. In support of convergent validity, the FSII-S scores correlated
of equivalence of the factor loadings across the sex and age
positively and significantly (p < .001) with measures of suicide
groups. For the next step (scalar invariance), model 3 included
risk - specifically the SBQ-R, BDI, and NA. Conversely, FSII-S
equivalence of means for items across groups and showed a
scores correlated negatively and significantly (p < .001) with
good fit, and variations fit index between models 3 and 2 did not
measures of suicide protection – specifically the FS, SWLS, and
exceed the cut-point. Following the invariance test, model 4 (strict
PA. The internal consistency coefficients are presented in Table 4;
invariance) showed a reasonable fit, and the comparison between
all values were satisfactory.
models 4 and 3 was less than the critical values across groups.
Therefore, the model was consistent in providing a robust test
Multi-Group Confirmatory Factor Analysis
regardless of sex and age in the adult population, which supported
A MG-CFA was used to determine whether there were gender the initial hypothesis of the FSII-S’s structural invariance across
or age group differences in the modelling of the FSII-S. The test sex and age groups.

Figure 1. Frequency distribution and information function of FSII-S

Table 4
Pearson Correlation Coefficients, Means (Standard Deviations), and internal consistency coefficients

1 2 3 4 5 6 7 Alpha Omega Guttman M(SD)

1.FSII-S – .89 .90 .88 7.87(3.88)

2.SBQ .60 – .72 .75 .71 5.95(2.77)
3.BDI .47 .51 – .90 .89 .89 28.98(8.07)
4.NA .41 .21 .42 – .84 .84 .83 22.57(7.60)
5.FS -.42 -.41 -.51 -.29 – .88 .88 .87 45.82(7.56)
6.SWLS -.35 -.38 -.47 -.26 .69 – .85 .86 .84 25.24(6.00)
7.PA -.23 -.21 -.41 -.14 .56 .50 – .84 .84 .85 34.28(6.99)

Note: N= 1013. FSII-S= Spanish Frequency of Suicidal Ideation Inventory; SBQ-R= Suicidal Behaviors Questionnaire Revised; BDI= Beck Depression Inventory; NA= Negative Affect Scale;
FS= Flourishing Scale; SWLS= Satisfaction with Life Scale; PA= Positive Affect Scale.
All correlations indexes were significant p< .001

257
 Nicolás Sánchez-Álvarez, Natalio Extremera Pacheco, Lourdes Rey, Edward C. Chang, and Olivia D. Chang

Table 5
Tests for invariance of FSII-S

Model X2 df CFI RMSEA SRMR Comparison ΔCFI ΔRMSEA ΔSRMR

Males vs Females
Model 1.Configural 1.368 10 .999 .001 .018 – – – –
Model 2.Metric 4.424 15 .999 .001 .032 2 vs 1 .000 .000 .014
Model 3.Scalar 5.321 19 .999 .001 .029 3 vs 2 .000 .000 .003
Model 4.Strict 6.385 24 .999 .001 .032 4 vs 3 .000 .000 .003
Young adults vs Middle adults vs Older adults
Model 1.Configural 2.266 15 .999 .001 .054 – – – –
Model 2.Metric 22.932 25 .999 .001 .075 2 vs 1 .000 .000 .021
Model 3.Scalar 34.336 33 .999 .011 .070 3 vs 2 .000 .010 .005
Model 4.Strict 52.167 43 .993 .025 .097 4 vs 3 .006 .014 .027

Discussion compared, not only across gender, but also across this large age
range within adulthood.
The validation of measures to evaluate suicidal thoughts is Despite these promising findings, some limitations of the present
a priority in the research and prevention of suicidal behaviors study are worth noting. First, our findings should be considered
(Chang et al., 2017), especially in countries such as Spain, where as preliminary, because of the use of student’recruited sampling.
the prevalence of suicide has increased during the last decades. In Although this sampling technique is a valuable tool increasingly
this study, we sought to investigate the psychometric properties used in research (Wheeler, Shanine, Leon, & Whitman, 2014),
of the FSII-S across a relatively large sample of Spanish adults. further research using traditional sampling procedures should be
Furthermore, the confirmatory factor analysis demonstrated a carried out to support the findings of the present study. Second,
one-dimensional structure for the FSII-S, with excellent internal for a more complete validation process, it would be interesting in
consistency and reliability, in line with the original version and future work to provide data about the FSII-S’s test–retest reliability.
other adaptation studies (Chang et al., 2017; Chang & Chang, Third, the cross-sectional nature of our research implies that it is
2016). Moreover, regarding construct validity, the FSII-S impossible to draw conclusions regarding cause and effect with
showed significant and positive correlations with depressive respect to the relationship between the FSII-S and measured well-
symptomatology and negative affect (Beck et al., 1993), being outcomes. Also, our findings suggest certain age stability,
consistent with the general notion that negative mood might be but cross-sectional designs are less suited to infer developmental
related to suicidal ideation. Additionally, regarding convergent trends. Furthermore, prospective research is needed to examine
validity, a high correlation was found between FSII-S scores and the age invariance across different time points, and the predictive
SBQ-R scores. Finally, in line with the expectation that positive validity of the FSII-S in accounting for risky behaviors and factors
psychological factors may protect against suicidal ideation, associated with suicide attempts over time in Spanish adults.
FSII-S scores were found to be significantly and negatively linked Finally, we did not control for other personal dimensions that may
to life satisfaction, positive affect, and flourishing (Hirsch et al., be associated with suicidal ideation (e.g., cognitive intelligence,
2007). Overall, these results are consistent with past findings personality traits). Future studies would benefit from the inclusion
(Chang et al., 2017; Chang & Chang, 2016), suggesting that of personality traits, cognitive intelligence, social desirability
future interventions focused on reducing negative conditions and measures, as well as questions about socio-economic status (Hegerl
increasing positive conditions might benefit from efforts to reduce et al., 2019). At the same time, future studies should extend our
suicidal ideation in Spanish adults. Additionally, the analysis of findings with the inclusion of clinical samples to provide clinical
multi-group invariance revealed that configural, metric, and cut-off points of this instrument.
scalar invariance existed across gender and age, suggesting that Notwithstanding these limitations, our study was designed to
the Spanish FSII items, as well as the underlying latent factor, provide support for the validation of the Spanish version of the FSII
are interpreted similarly by females and males. That is, the same in Spanish adults. Our findings suggest that the FSII-S might be a
structure of the FSII-S causes equal associations of items with useful tool for researchers when studying risk factors of suicide
the same factor and an equality of strength in the relationship of in Spanish adults, especially when a short and reliable measure
the construct with each item was corroborated for both genders. of suicidal ideation frequency is desired. In sum, these results
These findings are in line with previous literature that found no are an important contribution to the literature on psychometric
differences in suicidal ideation between sexes (Fonseca-Pedrero characteristics of brief suicide ideation measures, suggesting
et al., 2018) and suggest that the FSII can be reliably applied to invariance across male and female groups as well as three adult age
Spanish female and male adults. Similarly, the results did not show groups. Thus, the Spanish FSII may prove useful in the detection
significant differences between the three age groups, maintaining a and prevention of warning signs and early symptoms of those at
factorial and metric factor invariance. Accordingly, these findings high risk of suicide among Spanish-speaking populations.
show the consistency of the FSII-S in different adult stages and
are of great relevance for the evaluation of suicidal ideation in Acknowledgements
young adults, middle-aged, and older adults, specifically for the
longitudinal tracking of participants. Therefore, consistent with This work was supported by Junta de Andalucia/FEDER funds
our findings, Spanish FSII scores might thus be meaningfully (UMA18-FEDERJA-147).

258
 Frequency of Suicidal Ideation Inventory: Psychometric Properties of the Spanish Version

References

Atienza, F., Pons, D., Balaguer, I., & Merita, M. (2000). Propiedades psychoeducation. Psicothema, 31(2), 107-113. https://doi.org/10.7334/
psicométricas de la Escala de Satisfacción con la Vida en adolescentes psicothema2018.247
[Psychometric properties of the Satisfaction with Life Scale in Fonseca-Pedrero, E., Inchausti, F., Pérez-Gutiérrez, L., Aritio-Solana,
adolescents]. Psicothema, 12(2), 314-319. Retrieved from https://www. R., Ortuño-Sierra, J., Sánchez-García, M.,… Pérez-Albéniz, A.
redalyc.org/articulo.oa?id=72712226 (2018). Suicidal ideation in a community-derived sample of Spanish
Batterham, P., Ftanou, M., Pirkis, J., Brewer, J., Mackinnon, A., & adolescents. Revista de Psiquiatría y Salud Mental (English edition),
Christensen, H. (2015). A systematic review and evaluation of measures 11(2), 76-85. https://doi.org/10.1016/j.rpsmen.2018.02.008
for suicidal ideation and behaviors in population-based research. Franklin, J., Ribeiro, J., Fox, K., Bentley, K., Kleiman, E., Huang, X.,…
Psychological Assessment, 27(2), 501-512. https://doi.org/10.1037/ Franklin, C. (2017). Risk factors for suicidal thoughts and behaviors:
pas0000053 A meta-analysis of 50 years of research. Psychological Bulletin, 1-47.
Beck, A., Brown, G., & Steer, R. (1997). Psychometric characteristics of https://doi.org/10.1037/bul0000084
the Scale for Suicide with psychiatric outpatients. Behaviour Research García-Nieto, R., Blasco-Fontecilla, H., Paz-Yepes, M., & Baca-García,
and Therapy, 35(11), 1039-1046. https://doi.org/10.1016/S0005- E. (2013). Traducción y validación de la Self-Injurious Thoughts
7967(97)00073-9 and Behaviors Interview en población española con conducta suicida
Beck, A., Kovacs, M., & Weissman, A. (1979). Assessment of suicidal [Translation and validation of the Self-Injurious Thoughts and Behaviors
intention: The Scale for Suicide Ideation. Journal of Consulting and Interview in Spanish population with suicidal behavior]. Revista de
Clinical Psychology, 47(2), 343-352. https://doi.org/10.1037/0022- Psiquiatria y Salud Mental, 6(3), 101-108. https://doi.org/10.1016/j.
006X.47.2.343 rpsm.2012.07.001
Beck, A., Steer, R., Beck, J., & Newman, C. (1993). Hopelessness, García-Nieto, R., Parra-Uribe, I., Palao, D., López-Castroman, J., Sáiz,
depression, suicidal ideation, and clinical diagnosis of depression. P., García-Portilla, M.,… Baca-García, E. (2012). Protocolo breve de
Suicide and Life-Threatening Behavior, 23(2), 139. https://doi. evaluación del suicidio: fiabilidad interexaminadores [Brief suicide
org/10.1111/j.1943-278X.1993.tb00378.x evaluation protocol: Inter-examiner reliability]. Revista de Psiquiatria y
Beck, A., Steer, R., & Garbin, M. (1988). Psychometric properties of the Salud Mental, 5(1), 24-36. https://doi.org/10.1016/j.rpsm.2011.10.001
Beck Depression Inventory: Twenty-five years of evaluation. Clinical Gili, M., Roca, M., Basu, S., McKee, M., & Stuckler, D. (2013). The
Psychological Review, 8, 77-100. https://doi.org/10.1016/0272- mental health risks of economic crisis in Spain: Evidence from primary
7358(88)90050-5 care centres, 2006 and 2010. European Journal of Public Health, 23(1),
Blasco, M., Vilagut, G., Almenara, J., Roca, M., Piqueras, J., Gabilondo, 103-108. https://doi.org/10.1093/eurpub/cks035
A.,… Vives, M. (2019). Suicidal thoughts and behaviors: Prevalence Goñi-Sarriés, A., Blanco, M., Azcárate, L., Peinado, R., & López-
and association with distal and proximal factors in Spanish university Goñi, J. (2018). Are previous suicide attempts a risk factor for
students. Suicide and Life-Threatening Behavior, 49(3), 881-898. completed suicide? Psicothema, 30(1), 33-38. https://doi.org/10.7334/
https://doi.org/10.1111/sltb.12491 psicothema2016.318
Chang, E. C., & Chang, O. D. (2016). Development of the Frequency of Hansen, L., & Pritchard, C. (2008). Consistency in suicide rates in twenty-
Suicidal Ideation Inventory: Evidence for the validity and reliability two developed countries by gender over time 1874-78, 1974-76, and
of a brief measure of suicidal ideation frequency in a college student 1998-2000. Archives of Suicide Research, 12(3), 251-262. https://doi.
population. Cognitive Therapy and Research, 40(4), 549-556. https:// org/10.1080/13811110802101153
doi.org/10.1007/s10608-016-9758-0 Hatch, S., & Dohrenwend, B. (2007). Distribution of traumatic and other
Chang, E. C., Chang, O. D., Martos, T., Sallay, V., Najarian, A., & Lee, stressful life events by race/ethnicity, gender, SES and age: A review
J. (2017). Validity of the Frequency of Suicidal Ideation Inventory in of the research. American Journal of Community Psychology, 40, 313-
Hungarian adults. Death Studies, 41(4), 236-240. https://doi.org/10.10 332. https://doi.org/10.1007/s10464-007-9134-z
80/07481187.2016.1244226 Hawton, K., & van-Heeringen, K. (2009). Suicide. The Lancet, 373(9672),
Chang, S., Stuckler, D., Yip, P., & Gunnell, D. (2013). Impact of 2008 1372-1381. https://doi.org/10.1016/S0140-6736(09)60372-X
global economic crisis on suicide: Time trend study in 54 countries. Hegerl, U., Maxwell, M., Harris, F., Koburger, N., Mergl, R., Székely,
BMJ, 347(7925), 1-15. https://doi.org/10.1136/bmj.f5239 A.,… Gusmão, R. (2019). Prevention of suicidal behaviour: Results
Chen, F. (2007). Sensitivity of goodness of fit indexes to lack of of a controlled community-based intervention study in four European
measurement invariance. Structural Equation Modeling, 14(3), 464- countries. PLoS ONE, 14(11), 1-26. https://doi.org/10.1371/journal.
504. https://doi.org/10.1080/10705510701301834 pone.0224602
Córdoba-Doña, J., San-Sebastián, M., Escolar-Pujolar, A., Martínez-Faure, Hendriks, V., & Blanken, P. (1992) Snowball sampling: Theoretical and
J., & Gustafsson, P. (2014). Economic crisis and suicidal behaviour: practical considerations. In Hendricks, V., Blanken, P., & Adriaans, N.
The role of unemployment, sex and age in Spain. International Journal (Eds.), Snowball sampling: A pilot study on cocaine use (pp. 17-35).
Form Equity in Health, 13(55). https://doi.org/10.1186/1475-9276-13- Rotterdam, the Netherlands: IVO.
55 Hirsch, J., Duberstein, P., Chapman, B., & Lyness, J. (2007). Positive affect
Demerouti, E., & Rispens, S. (2014). Improving the image of student- and suicide ideation in older adult primary care patients. Psychology
recruited samples: A commentary. Journal of Occupational and and Aging, 22(2), 380-385. https://doi.org/10.1080/10937404.2015.10
Organizational Psychology, 87(1), 34-41. https://doi.org/10.1111/ 51611.
joop.12048 IBM, S.P.S.S. (2015). IBM SPSS Statistics Version 23. International
Diener, E., Emmons, R., Larsen, J., & Griffin, S. (1985). The Satisfaction Business Machines Corp., Boston.
with Life Scale. Journal of Personality Assessmemt, 49(1), 71-75. Joiner, T., Rudd, M., & Rajab, M. (1997). The modified scale for suicidal
https://doi.org/10.1207/s15327752jpa4901_13 ideation: Factors of suicidality and their relation to clinical and
Diener, E., Wirtz, D., Tov, W., Kim-Prieto, C., Choi, D., Oishi, S., & diagnostic variables. Journal of Abnormal Psychology, 106(2), 260-
Biswas-Diener, R. (2010). New well-being measures: Short scales to 265. https://doi.org/10.1037/0021-843X.106.2.260
assess flourishing and positive and negative feelings. Social Indicators Kline, R. (2015). Principles and practice of structural equation modeling.
Research, 97(2), 143-156. https://doi.org/10.1007/s11205-009-9493-y Guilford publications.
Extremera, N., Quintana-Orts, C., Mérida-López, S., & Rey, L. (2018). Liang, J., & Bentler, P. (2004). An EM algorithm for fitting two-level
Cyberbullying victimization, self-esteem and suicidal ideation in structural equation models. Psychometrika, 69(1), 101-122. https://doi.
adolescence: Does emotional intelligence play a buffering role? Frontiers org/10.1007/BF02295842
in Psychology, 9, 367 https://doi.org/10.3389/fpsyg.2018.00367 López-Bernal, J., Gasparrini, A., Artundo, C., & McKee, M. (2013).
Fernández-Artamendi, S., Al-Halabí, S., Burón, P., Rodríguez- The effect of the late 2000s financial crisis on suicides in Spain: An
Revuelta, J., Garrido, M., González-Blanco, L.,… Bobes, J. (2019). interrupted time-series analysis. European Journal of Public Health,
Prevention of recurrent suicidal behavior: Case management and 23(5), 732-736. https://doi.org/10.1093/eurpub/ckt083

259
 Nicolás Sánchez-Álvarez, Natalio Extremera Pacheco, Lourdes Rey, Edward C. Chang, and Olivia D. Chang

Muñiz, J., & Fonseca-Pedrero, E. (2019). Ten steps for test development. Sanz, J., & Vázquez, C. (1998). Fiabilidad, validez y datos normativos del
Psicothema, 31(1), 7-16. https://doi.org/10.7334/psicothema2018.291 Inventario para la Depresión de Beck [Reliability, validity and regulatory
Osman, A., Bagge, C., Gutiérrez, P., Konick, L., Kopper, B., & Barrios, inventory data for the Beck Depression Inventory]. Psicothema, 10(2),
F. (2001). The Suicidal Behaviors Questionnaire-Revised (SBQ-R): 303-318. https://doi.org/7467-13170-1
Validation with clinical and nonclinical samples. Assessment, 8(4), 443- Soto-Sanz, V., Piqueras, J., Rodríguez-Marín, J., Pérez-Vázquez, M.,
454. https://doi.org/10.1177/107319110100800409 Rodríguez-Jiménez, T., Castellví, P.,… Alonso, J. (2019). Self-esteem
Ramírez-Maestre, C., Correa, M., Rivas, T., López-Martínez, A., Serrano- and suicidal behaviour in youth: A meta-analysis of longitudinal
Ibáñez, E., & Esteve, R. (2017). Psychometric characteristics of the studies. Psicothema, 31(3), 246-254. https://doi.org/10.7334/
Flourishing Scale-Spanish Version (FS-SV). The factorial structure psicothema2018.339
in two samples: Students and patients with chronic pain. Personality Timmons, A. (2010). Establishing factorial invariance for multiple-group
and Individual Differences, 117, 30-36. https://doi.org/10.1016/j. confirmatory factor analysis. KUant Guide, 22. Retrieved from http://
paid.2017.05.035 crmda.dept.ku.edu/resources/kuantguides/22.Factorial_Invariance_
Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Guide.pdf
Journal of Statistical Software, 48(2), 1-36. Retrieved from https:// Watson, D., Clark, L., & Tellegen, A. (1988). Development and validation
econpapers.repec.org/article/jssjstsof/v_3a048_3ai02.htm%0Ahttp:// of brief measures of positive and negative affect: The PANAS scales.
www.jstatsoft.org/v48/i02/ Journal of Personality and Social Psychology, 54(6), 1063-1070.
Rubio, G., Montero, I., Jáuregui, J., Salvador, M., Marín, J., & Carrasco, https://doi.org/10.1037/0022-3514.54.6.1063
J. (1998). Validación de la Escala de Riesgo de Violencia de Plutchik Wheeler, A., Shanine, K., Leon, M., & Whitman, M. (2014). Student-
en población española [Validation of the Plutchik Violence Risk Scale recruited samples in organizational research: A review, analysis,
in the Spanish population]. Archivos de Neurobiología: Revista de and guidelines for future research. Journal of Occupational and
Psiquiatría y Disciplinas Afines, 61(4), 307-316. Organizational Psychology, 87(1), 1-26. https://doi.org/10.1111/
Sandín, B. (2003). Escalas Panas de afecto positivo y negativo para niños joop.12042
y adolescentes (PANASN) [Panas scales of positive and negative affect World Health Organization (2018). Suicide. Retrieved from http://www.
for children and adolescents (PANASN)]. Revista de Psicopatología who.int/mediacentre/factsheets/fs398/en/
y Psicología Clínica, 8(2), 173-182. https://doi.org/10.5944/rppc. Zenisky, A., & Hambleton, R. (2015). A model and good practices for score
vol.8.num.2.2003.3953 reporting. In Handbook of test development (pp. 601-618). Routledge.

260