Adapting the ryff scales of psychological well being a 28 item vietnamese version for university students

176 Adapting the Ryff Scales of Psychological Well-being: a Item Vietnamese Version for University Students 28-Raul Calderon Jr*, Nguyen Thi Nga, Truong Quang Tien, Bui Thi Tu Quyen, N

176

Adapting the Ryff Scales of Psychological Well-being: a Item Vietnamese Version for University Students

28-Raul Calderon Jr*, Nguyen Thi Nga, Truong Quang Tien, Bui Thi Tu Quyen,

Nguyen Hoang Minh Thuan, Vu Viet Bao

Abstract: Research in psychological well-being and correlates to general health have blossomed

in the last 20-25 years However, little research exists in health and psychological well-being of the Vietnamese population, partially due to a lack of valid and reliable research instruments available in the Vietnamese language This study aimed to produce a short, valid and reliable research instrument in Vietnamese to assess the psychological well-being of university students The 54-item Ryff Scales of Psychological Well-being (RSPWB), a widely used and validated instrument, was the instrument of choice A research questionnaire, which included the RSPWB, was translated and completed by 253 Vietnamese undergraduate students (18-27 years of age) Analyses included the use of partial least squares, structural equation modeling (PLS-SEM) - a second generation statistical technique for testing complex path models, to evaluate the validity and reliability of the 54 items As a result of the research, the PLS-SEM analyses produced a multidimensional model with six factors (autonomy, environmental mastery, purpose in life, personal growth, positive relations with others, and self-acceptance) and 28 items that demonstrated good convergent and discriminant validity and reliability Moreover, the 28-item Vietnamese version of the RSPWB supported the same multidimensional six-factor structure as proposed in the original RSPWB In conclusion, the adapted 28-item Vietnamese version of the RSPWB is shorter in length, has good psychometric properties, and helps to fill a current gap of available research instruments for advancing the current understanding of psychological well-being and relationships to health in Vietnamese university students

Keywords: Ryff Scales of Psychological Well-being; Vietnam; University Students;

Validation; PLS-SEM

Received: 6 th January 2019 ; Revised 15 th April 2019; Accepted 23 rd April 2019

DOI: https://doi.org/10.33100/jossh5.2.Calderon.etal

1 Introduction

Globalization and technology have

increased access to information and

 Raul Calderon - Faculty of Sports Science, Kasetsart

University, Kamphaeng Saen Campus, Nakhon Pathom,

Thailand, email: raulthedoc@yahoo.com

Nguyen Thi Nga, Truong Quang Tien, Bui Thi To

Quyen - Hanoi University of Public Health, Vietnam

Nguyen Hoang Minh Thuan, Vu Viet Bao - Ho Chi Minh

City University of Sports, Vietnam

scientific materials to researchers around the world Researchers from most countries can access a variety of reliable and valid research instruments through the worldwide web Although access to scientific materials has become easier, it does not mean that instruments or scales in one language can automatically be used in countries with different languages and/or cultures Sousa & Rojjanasrirat (2011) recommend a series of steps in the translation and cross-validation

process before instruments and scales are

used in languages and/or cultures other than

their original language Implementing such

steps can improve the overall quality of

cross-cultural research

2 Background and significance

Positive psychology has flourished over

the last 20-25 years There is now more

attention given to the positive aspects of

human development and functioning

(Seligman 2012) One of the specific

domains that has garnered much attention is

the field of well-being and in particular

psychological well-being (Ryff 2013)

Generally, well-being is thought to be a

multi-faceted concept that includes

subjective, social, and psychological

dimensions as well as health-related

behaviors (Seifert 2005) Research into the

general field of well-being has advanced

over the years and now is most prominent in

the field of empirical psychology The

amount of literature in the area of well-being

(including the two main branches of

subjective and psychological well-being)

has flourished as researchers investigate this

complex construct that concerns the

understanding of optimal human experience

and functioning (Deci and Ryan 2006)

In the area of subjective well-being, the

work by Diener (Diener et al 1999; Diener

E 1984) into the nature of subjective

well-being and life satisfaction, Lyubomirsky

(1998) into subjective happiness, and Deci

(1990) into personality integration, to

mention only a few, have helped advance

the understanding of the general field of

subjective well-being However, one

criticism of subjective well-being research

has been that it is not fully grounded in

theory

Ryff (1989) addressed this criticism by synthesizing the work of leading theorists such as Bühler‟s (1935) basic life tendencies that work toward fulfillment of life, Allport‟s (1961) conception of maturity, Neugarten‟s (1968) descriptions of personality change in adulthood and old age, Jahoda‟s (1979) positive criteria of mental health, Erikson‟s (1994) psychosocial stage model, Jung‟s (1935) formulation of individuation, Rodger‟s (1961) view of the fully functioning person, and Maslow‟s (1962) conception of self-actualization, and proposed a theory-based multidimensional model of psychological well-being (PWB) The theory-based multidimensional model proposed by Ryff (1989) consists of six dimensions The theoretical dimensions are (1) autonomy, (2) environmental mastery, (3) purpose in life, (4) personal growth, (5) positive relations with others, and (6) self-acceptance These six dimensions taken together are thought to compose the theoretical construct of PWB

In other words, PWB is a composite construct consisting of multiple sub-constructs Each sub-construct contributes a unique and important aspect of the broader construct of PWB

To measure these constructs, Ryff developed three versions of the Ryff Scales

of Psychological Well-being (RSPWB), including 84, 54, and 18 item versions All three versions have been used, validated, and sometimes modified to five to seven items per construct in previous studies In general, the longer versions show stronger psychometric properties, thus are recommended for serious research projects (van Dierendonck 2004; Ryff 1989; Ryff and Keyes 1995)

The RSPWB is considered to be one of the leading instruments for assessing PWB Since the development of the RSPWB, it has

been used in over 350 studies and translated

into numerous languages (Cheng & Chan

2005; Freire et al 2016; Kitamura et al

2004; Lapa 2015; Ryff 2013; Woo Kyeong

2013) However, the majority of the

research using the RSPWB is mostly from

English speaking and western populations

Therefore, a significant amount of the

current knowledge and understanding of

PWB primarily describes the US and

European populations Less is known about

PWB in Asian populations since few Asian

countries have utilized validated instruments

to assess PWB

To assess PWB in an Asian population

using the RSPWB, the instrument must be

translated and validated into the host

language The first published study in

Vietnam to adopt the RSPWB for a

Vietnamese population was conducted by

Dang (2017) Her study indicated that the

84-item RSPWB (14-items/construct) had

good reliability However, this study was

not without its limitations First, the study‟s

translation method was not explained

thoroughly Second, convergent and

discriminant validity was not fully

established Third, the 84-item version is

long and takes the longest time to complete

Fourth, the second-order construct of PWB was not tested for significance Fifth, some

of the sub-construct items overlapped with different items from other sub-constructs

To confirm the findings by Dang (2017) and to produce a shorter user-friendly version of the RSPWB for the Vietnamese population the following study was conducted with the specific aims:

(1) To use a reliable, accepted method of translation for the RSPWB

(2) To determine if a fewer number of items of the RSPWB would produce acceptable convergent and discriminant validity and reliability using partial least squares, structural equation modeling (PLS-SEM) - a second generation statistical technique

(3) To determine the path coefficients of the six sub-constructs (autonomy, environmental mastery, purpose in life, personal growth, positive relations with others, and self-acceptance) and test the significance of the second-order construct of PWB

(4) To compare the factorial structure of the adapted Vietnamese RSPWB to the original factorial structure of the RSPWB

3 Method

The three stages of this study are depicted in Table 1

Table 1: List of components for each stage in the study

Stage 1

Translation of research questionnaire

Explored students‟ understanding of

PWB and its related sub-constructs

Components of the Vietnamese

RSPWB questionnaire

Conducted pilot study

Stage 2 Participants in the main study

Data collection

Stage 3 Statistical analysis Sample size and power

Ethical approval

Stage 1

Translation of Vietnamese RSPWB

questionnaire

A collaborative and iterative method was

used for translating the study questionnaire

into Vietnamese Although this method does

not use a back translation step, it is still an

acceptable alternative translation method

This method provided a conceptual

equivalence of translated items, which

benefits comprehension and meaning to the

respondent (Douglas and Craig 2007)

Additionally, throughout the iterative steps,

the translators referred back to the original

English version when making edits

A group of five native Vietnamese

speakers, one Vietnamese university English

lecturer (professional translator), and two

Vietnamese research university professors,

all who were fluent in English, were

recruited to help translate the English

questionnaire into Vietnamese The

collaborative and iterative process

proceeded as follows:

Each translator was assigned

approximately one-fifth of the English

questionnaire to translate All translated

parts were combined to create a complete

translated Vietnamese questionnaire draft

The translated questionnaire draft was

then reviewed and rated by each translator

for accuracy using a scale from 0 (no

accuracy) to 100 (excellent accuracy) Any

item that did not receive an accuracy rating

of 90 or higher was flagged In other words,

every item in the questionnaire was checked

for accuracy/equivalency to English version

by five independent translators This

assured a high-quality translation because

each item needed the agreement of all

translators to remain in the questionnaire

Items that were flagged (i.e., received an

accuracy rating of less than 90 by any of the

translators) were sent back to the translators for review and re-translation After the flagged items were corrected, a new edited draft was created

The edited draft of the questionnaire was reviewed for accuracy/equivalency and edited by the professional translator (university English class lecturer) This provided an additional check on the quality

of the translations done by the first set of translators Once this step was completed, a final draft was created

Lastly, the final draft was reviewed and edited by two research university professors familiar with Vietnamese culture and research surveys

This collaborative and iterative process generated the final version of the questionnaire

Explored students’ understanding of PWB and its related sub-constructs

Very little research of the PWB of the Vietnamese population has been done; thus,

as a first step, a focus group session was scheduled with a group of university students Participants in the focus group included five female and four male undergraduate students from the English club at the Vietnam Maritime University of Haiphong All students were proficient in English, and the session was conducted in English by the lead author The session lasted approximately 90 minutes The main focus of the session was to explore the student‟s understanding of the concept of psychological well-being and its related sub-constructs of autonomy, environmental mastery, personal growth, purpose in life, positive relations with others, and self-acceptance Students were asked questions

of how they perceived PWB and its related sub-constructs In general, the students had

an adequate grasp of these constructs and

could relate to how the constructs related to

their lives and the lives of other students

The overall impression derived from the

focus group was that students were familiar

and comfortable in speaking about

well-being and related constructs

Components of Vietnamese RSPWB

questionnaire

The Vietnamese RSPWB questionnaire

included the 54-item version of the RSPWB

and 15 demographic questions

The original RSPWB were developed to

measure the multidimensional nature of

psychological well-being (Ryff 1989) There

are three versions of the RSPWB, with 84,

54, and 18 items per version, with 14, 9, and

3 items per construct, respectively For this

study, the version with 54-items was

selected in order to minimize the overall

length of the questionnaire and lessen

survey fatigue The 54-item version of the

RSPWB has six sub-scales (9-items per

scale) The scales are (1) Autonomy

(self-determination and independence), (2)

Environmental Mastery (a sense of mastery

and competence in managing one‟s

environment), (3) Personal Growth (a

feeling of continued development), (4)

Purpose in Life (having goals in life and a

sense of direction), (5) Positive Relations

with Others (having warm, satisfying,

trusting relationships with others), and (6)

Self-acceptance (possessing a positive

attitude towards the self) (see Supplemental

Material 2) The six sub-scales are

hypothesized to form the theoretical

construct of PWB and changes in the six

sub-scales will cause changes in the

construct of PWB Each sub-scale consists

of a mix of positive and negative sentences

Participants responded using a six-point

Likert scale ranging from 1 (strongly

disagree) to 6 (strongly agree) Responses

to negatively scored items (-) are reversed in

the final scoring procedures so that high scores indicate high self-ratings on the dimension assessed Internal consistency for the different sub-scales in the English RSPWB version range from 86 to 93 and show good validity (van Dierendonck 2004) Dang (2017) reported internal consistency ranging from 83 to 92 in the Vietnamese version of the RSPWB

Conducted pilot study

Students from the English club at the Vietnam Maritime University of Haiphong sent invitations (via social media) to their university friends asking them to complete the Vietnamese RSPWB questionnaire online Thirty-nine undergraduate students (17 female, and 22 male) responded to the invitation After 3-4 days, these 39 students were asked to complete the questionnaire a second time Twenty-three of the 39 students completed the questionnaire a second time Most students completed the questionnaire within 20-25 minutes Participants used their smartphones, at a convenient time of their choosing, to access

a link sent to them via social media directing them to the questionnaire online hosted on SurveyMonkey, a commercial data collection website (Survey n.d.) Reliability analysis of the six sub-constructs yielded good internal consistency ranging from 69

to 90 Test re-test reliability ranged from 0.849 (positive relations with others) to

0.523 (self-acceptance)

Stage 2

Participants in the main study

The participants of the main study were public health undergraduate students enrolled at the Hanoi University of Public Health, Vietnam A total of 530 public health students (77.3% female and 22.7% male) were enrolled in the university during

the school year 2016-17) Research team

members contacted the Department of

Undergraduate Education, the Department

of Student Affairs, and all undergraduate

course instructors (total of 16 instructors) to

explain the purpose of the study All course

instructors (four instructors per grade level,

1st, 2nd, 3rd, and 4th year) agreed to allow

their classes to participate in the study

Data collection

A population sampling method was used

for this study All students enrolled in the

university were eligible and invited, via

individual classrooms, to complete the

Vietnamese RSPWB questionnaire by

accessing the questionnaire online using

their smartphones or computer A research

team member visited each classroom to

explain the purpose of the study, invite

students to participate, and provide the link

to the online questionnaire Data collection

via the SurveyMonkey website was open

approximately five weeks (during the

Fall-semester of 2016) The researchers sent

weekly messages, during the open data

collection period, to course instructors

asking them to remind and encourage their

students to complete the online

questionnaire Researchers also enlisted the

help of school clubs, posted reminder

posters around campus, and used social

media to remind and encourage students to

participate Participation was voluntary

Students did not receive any course credit

for completing the questionnaire A total of

253 undergraduate students (81% female,

and 19% male, with a median age of 20)

completed the Vietnamese RSPWB

questionnaire for a 45.7% response rate Ten

cases were deleted due to missing data

(>10%) leaving a total N of 243 for analysis

Less than 5% of the cases had RSPWB

The SmartPLS version 2.0 software (Ringle, Wende, and Will 2005) was used to determine the reliability and validity of the PWB measurement and structural model The PLS-SEM statistical technique was chosen for three important reasons: (1) PLS-SEM can provide unique theoretical insights and prevent the misspecification of statistical models and hence erroneous results; (2) PLS-SEM is the preferable statistical technique when a model includes both reflective and formative constructs; and (3) PLS-SEM should be used when a model includes formative factors and has more than 40-50 indicators (Lowry and Gaskin 2014)

Lowry and Gaskin (2014) recommend using a 4-step approach for PLS-SEM analysis

Step 1: model specification Figure 1 shows the specified model that was tested The figure shows the measurement model, which included six reflective first-order latent variables (LVs): (1) autonomy, (2) environmental mastery, (3) purpose in life, (4) personal growth, (5) positive relations with others, and (6) self-acceptance and the structural model, which included one second-order formative LV, Psychological Well-being Each first-order LV had nine indicators, which were hypothesized to reflect the theoretical construct of their corresponding LV The direction of the indicator arrows is important as it determines whether the LV should be specified as a reflective or formative

construct Reflective construct arrows point

from the construct (causal) to their

indicators (effect) In other words, a change

in the construct will cause a change in the

indicators On the other hand, formative

construct arrows point in the opposite

direction, from the indicators (causal) to an

LV (effect) So, a change in the indicators

will cause a change in the LV PWB is a second-order formative construct because the arrows of the six LVs all point to it Specifying reflective and formative LVs, as mentioned above, was important to prevent the misspecification of the statistical model and hence avoid erroneous results

Figure 1: Measurement and Structural Model of Psychological Well-being and proposed six sub-constructs:

Autonomy, Environmental Mastery, Personal Growth, Purpose in Life, Positive Relations with Others, and

Self-Acceptance

Step 2: determine the construct validity

of reflective constructs The SmartPLS2

software was used to establish convergent

and discriminant validity To run the

required factor analysis the „factor

weighting‟ scheme and bootstrapping

procedure of „5000 sub-samples‟ were

selected The PLS algorithm default report

table (Bootstrapping, Outer loadings (Mean,

STDEV, T-Values)) was used to determine

convergent validity, and the cross-loadings

and average variance extracted (AVE) tables

were used to determine discriminant

validity An example of using SmartPLS2 to

perform a factor analysis can be seen on

Gaskination‟s YouTube channel (Gaskin

2012)

Step 3: establish the reliability of the

reflective constructs: Reliability is a

measure of internal consistency required in

reflective (internally correlated) latent

variables To establish reliability, the

composite reliability measures should be

greater than 0.700 (Chin 1998) The PLS

algorithm computes a composite reliability

score, which is found in the default report

under the overview tab of the quality criteria

section When using PLS-SEM composite

reliability is generally regarded as the more

appropriate criterion to establish internal

consistency reliability as compared to

Cronbach‟s Alpha (Hair et al 2012) The

PLS algorithm (selecting „factor weighting‟

scheme) and bootstrapping procedure

(selecting „5000 sub-samples‟) was used for

determining composite reliability

Step 4: provide and interpret final

statistics: In the last step, the measurement

model statistics (Table 6) and a summary of

the path coefficients and significance levels

of the structural model (Table 7) were

calculated The PLS algorithm (selecting

„path weighting‟ scheme) and bootstrapping

procedure (selecting „5000 sub-samples‟) were used to estimate the structural model

Sample size and power

Sample size in PLS is commonly determined by either multiplying 10 times the scale with the largest number of formative indicators or by multiplying 10 times the largest number of structural paths directed at a particular construct in the structural model (Lowry and Gaskin 2014) This study had six structural paths (see figure 1) directed at one second-order construct, which means the study needed a minimum sample size of 60 (10 * 6) However, this method has been criticized for being too liberal To ensure the study had enough power over 250 participants (more than four times the minimum required) were recruited

Ethical approval

Approval for this study was granted by the university research review board Students received written informed consent

at the beginning of the online questionnaire and were informed that all data collected would be anonymous

4 Results

A translated and easy to use formatted Vietnamese version of the 54-item RSPWB was used to collect the data (see supplemental material 1) At the end of the questionnaire, the resulting shorter version with 28 items is indicated by item number Tables 2-6 present the results related to Aim #2: To determine if a fewer number of items of the RSPWB would produce acceptable convergent and discriminant validity and reliability

Table 2 provides evidence of the significance of the loadings of each

indicator within each latent variable All

indicators of the latent variables should have

a significant t-statistic (> 1.96) in order to

demonstrate adequate convergent validity

As shown in Table 2, the data meets this

criterion, except items aut3, pg14, pl11, and

sa12 The four non-significant (ns)

indicators were removed from the model and not included in further analysis Overall,

50 of the 54 indicators among the six latent variables had significant t-statistic values This suggests that 50 indicators were valid and converged onto their respective construct

Table 2: T-statistics for Convergent Validity

Construct (latent variable) Indicator t-statistic p value

After providing evidence of convergent

validity in Table 2, discriminant validity was

evaluated using two separate statistical

outputs from the PLS algorithm,

cross-loadings and AVE tables, respectively

Table 3 shows the final matrix of

loadings and cross-loadings for the

reflective indicators in the model The

loadings of the indicators in this table should

be greater for the LV to which they

theoretically belong than for any other LV

Discriminant validity is adequate if the

cross-loadings (with other LVs) are more

than the absolute value of 0.100 distant from

the loading on the primary LV (Gefen and

Straub 2005) Through a repeated step

process, indicators that had loadings of <0.4

on the primary LV (Hair Jr et al 2016) or

cross-loadings less than the absolute value

of 0.100 distant from the loading on the primary latent variable were dropped from the model one at a time After each run of the PLS algorithm (i.e., dropping of low loading indicators and removal of conflicting cross-loadings) the effect on remaining LV loadings and cross-loading were checked for changes Removing items

in this manner actually improved discriminant validity since low loadings and cross-loadings not meeting stated criteria generally brought down the average loading for the LV The end result of this process resulted in a final adapted model consisting

of 28 indicators (i.e., Aut = 5, Em = 4, Pg =

5, Pl = 4, Pr = 5, and Sa = 5), which suggests these 28 indicators had adequate convergent and discriminant validity

Table 3: Loadings of the Measurement Items

Autonomy Environmental

Mastery

Personal Growth

Purpose in Life

Positive Relation with Others

Acceptance