This article was downloaded by: [Vivian W. Q. Lou]On: 07 January 2014, At: 15:30Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Aging & Mental HealthPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/camh20

Factorial structure and psychometric properties ofa brief version of the Reminiscence Functions Scalewith Chinese older adultsVivian W.Q. Loua & Jacky C.P. Choya

a Department of Social Work and Social Administration, Sau Po Centre on Ageing, TheUniversity of Hong Kong, Pokfulam, Hong KongPublished online: 03 Dec 2013.

To cite this article: Vivian W.Q. Lou & Jacky C.P. Choy , Aging & Mental Health (2013): Factorial structure andpsychometric properties of a brief version of the Reminiscence Functions Scale with Chinese older adults, Aging & MentalHealth, DOI: 10.1080/13607863.2013.860423

To link to this article: http://dx.doi.org/10.1080/13607863.2013.860423

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Factorial structure and psychometric properties of a brief version of the Reminiscence

Functions Scale with Chinese older adults

Vivian W.Q. Lou* and Jacky C.P. Choy

Department of Social Work and Social Administration, Sau Po Centre on Ageing, The University of Hong Kong, Pokfulam, Hong Kong

(Received 1 February 2013; accepted 16 October 2013)

Objective: The current study aims to examine the factorial structure and psychometric properties of a brief version of theReminiscence Functions Scale (RFS), a 14-item assessment tool of reminiscence functions, with Chinese older adults.Method: The scale, covering four reminiscence functions (boredom reduction, bitterness revival, problem solving, andidentity) was translated from English into Chinese and administered to older adults (N ¼ 675). Confirmatory factoranalysis and hierarchical confirmatory factor analysis were conducted to examine its factorial structure, and itspsychometric properties and criterion validity were examined.Results: Confirmatory factor analysis supports a second-order model comprising one second-order factor and four first-order factors of RFS. The Cronbach’s alpha of the subscales ranged from 0.75 to 0.90.Conclusion: The brief RFS contains a second-order factorial structure. Its psychometric properties support it as a soundinstrument for measuring reminiscence functions among Chinese older adults.

Keywords: reminiscence; assessment; factorial structure; psychometric properties; Chinese older adults

Introduction

Reminiscence is the universal process of recalling one’s

past memories, which is a naturally occurring phenome-

non throughout the lifespan (Bluck & Levine, 1998;

Westerhof, Bohlmeijer, & Webster, 2010). Reminiscing

is associated with happiness (Webster, 1998; Webster &

McCall, 1999), state and trait anxiety (Cully, LaVoie, &

Gfeller, 2001), meaning in life (Cappeliez & O’Rourke,

2002), and life satisfaction and psychological distress

(Cappeliez, O’Rourke, & Chaudhury, 2005). Previous

studies have recognized that reminiscence can serve dif-

ferent functions, which are referred to as the taxonomy

of reminiscence (Coleman, 1986; Lo Gerfo, 1980; Wong

& Watt, 1991). In order to assess the functions of remi-

niscence, the English version of the Reminiscence Func-

tions Scale (RFS) was first developed in the 1990s

(Webster, 1993, 1997), and then validated in French

(Mezred, Petigenet, Fort, Blaison, & Gana, 2006) and

Portuguese (Goncalves, Guedes, Fonseca, & Martin,

2010).

Although cross-cultural studies have revealed that

Chinese and Americans demonstrate differences in mem-

ory content (i.e. relation-centered vs. self-focused auto-

biographical history; Conway, Wang, Hanyu, & Haque,

2005), we argue that the functions served by reminiscence

should be similar across different cultural contexts. Given

the large number of Chinese people worldwide, this study

aims to examine the factorial structure and psychometric

properties of the brief version of the RFS with Chinese

older adults, thereby generating a better understanding of

reminiscence functions in research and social practices

among the Chinese population.

Core components of reminiscence and its measure

Webster (1993, 1997) empirically identified eight uses of

reminiscence and, on that basis, constructed the 43-item

RFS, the most widely adopted assessment tool to date for

measuring reminiscence functions (Westerhof et al.,

2010). Robitaille and his colleagues (2010) recently con-

firmed the original eight-factor structure, yet questioned

the value of some items. Accordingly, the scale was modi-

fied by including 29 items for the eight factors. Adequate

levels of internal consistency of the subscales, ranging

from 0.76 to 0.86, supported this modified version of the

RFS. Test–retest reliability, at an 8-month interval, ranged

from 0.48 to 0.63.

Among the eight reminiscence functions identified,

four of them (boredom reduction, bitterness revival, prob-

lem solving, and identity) have demonstrated strong asso-

ciations with mental health in studies exploring the

relationship between reminiscence and psychological

well-being (Cappeliez & O’Rourke, 2006; Cully et al.,

2001; O’Rourke, Cappeliez, & Claxton, 2011; Webster,

1998). Boredom reduction refers to recalling past experi-

ences to escape from a current understimulating environ-

ment. Bitterness revival involves ruminating on previous

memories of painful life events and unjust treatments.

Problem solving means using past coping strategies to

solve current problems. Finally, identity refers to search-

ing for the meaning of life and a sense of who we are

(Webster, 1997). Cappeliez and O’Rourke (2006) devel-

oped a model explaining the relationship between the

reminiscence functions and well-being in later life (i.e.

life satisfaction, psychiatric distress, perceived health,

and health conditions). This model proposed that four

*Corresponding author. Email: wlou@hku.hk

� 2013 Taylor & Francis

Aging & Mental Health, 2013

http://dx.doi.org/10.1080/13607863.2013.860423

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14

functions of reminiscences impact on subjective well-

being via two possible mechanisms. One mechanism

enhances self-positive evaluation via problem solving

and identity functions, whereas the second mechanism

triggers self-negative evaluation via boredom reduction

and bitterness revival (O’Rourke et al., 2011). Moreover,

compared to other reminiscence functions, previous stud-

ies on the association of RFS subscales with mental

health have shown a more consistent relationship

between these four functions and psychological well-

being (Cappeliez & O’Rourke, 2006; Cappeliez et al.,

2005; Cully et al., 2001; Korte, Bohlmeijer, Westerhof,

& Pot, 2011; Westerhof et al., 2010; Wong & Watt,

1991). Hence, the present study focused on the brief

RFS, which comprised four reminiscence functions,

namely boredom reduction, bitterness revival, problem

solving, and identity.

However, the factorial structure and psychometric

properties of the brief RFS were subjected to re-evalua-

tion. When the RFS was initially developed, the problem

solving and identity items loaded onto the same factor in

exploratory factor analysis (EFA; Webster, 1993). How-

ever, in a replication of this original study, Webster

(1997) reported results supporting the distinction of the

two factors. Yet, in a recent study (Robitaille, Cappeliez,

Coulombe, & Webster, 2010), EFA of the RFS again

reflected that the two components loaded onto the same

factor. More empirical evidence is needed in order to the-

oretically regard problem solving and identity as distinct

notions.

The factorial structure of the brief RFS was exam-

ined using the original English version, and also vali-

dated in French and Portuguese (Goncalves et al., 2010;

Mezred et al., 2006). Importantly, previous studies have

also used a total composite score to assess the reminis-

cence functions (Cully et al., 2001). Is it possible that

reminiscence functions could be represented according

to a one-factor structure? If so, is this one-factor struc-

ture valid at the levels of direct observation and/or sec-

ond-order? Answers to these questions would not only

contribute to theories concerning reminiscence func-

tions, but also offer practical implications for clinical

interventions.

In sum, prior studies investigating the factorial struc-

ture of RFS have revealed that its factorial structure

deserves further examination. Specifically, the possibility

of a distinction between problem solving and identify fac-

tors must be examined. Therefore, the first objective of

the present study was to examine model fit by comparing

a three-factor model and a four-factor model of the brief

RFS. Moreover, considering both factor scores and total

sum scores have often been used to examine relationships

between reminiscence and psychological well-being, it is

crucial to examine the underlying structure of such an

aggregation. Therefore, we examined the one-factor

model fit at the direct observation and second-order factor

levels. Finally, psychometric properties of the brief RFS

were examined among Chinese older adults, a previously

unexplored population.

Design and methods

Participants

The participants were older adults (N ¼ 675), 60 years or

older, who either lived alone or only with their spouse in

the community. They comprised 227 men (33.6%) and

448 women (66.4%), and their ages ranged from 60 to

100 years (M ¼ 77.6, SD ¼ 7.0). With regard to the par-

ticipants’ levels of education, 42.4% had received no pre-

vious schooling, 43.7% had completed primary education,

and 13.9% had completed secondary education.

Procedure

The current study was part of a larger study examining the

psychological well-being of community-dwelling older

adults in Hong Kong. All participants were recruited from

a local district elderly community center (DECC). DECCs

are community centers operated by non-government

organizations and accountable for the Hong Kong

Special-Administrative Region Government. Located in

every district in Hong Kong, they provide comprehensive

support services to meet elders’ community needs. Their

scope of service includes health education, community

reach and networking, career support services, social and

recreational services, and case management. Each DECC

serves a similar function and accommodates a similar

population of older adults in Hong Kong.

Four reminiscence functions, namely boredom reduc-

tion, bitterness revival, problem solving, and identity,

were selected deliberately for examination in the present

study. Only the items of these functions from the modified

RFS (RFS-29; Robitaille et al., 2010) were translated into

Chinese by the principal investigator and a research assis-

tant with a psychology background. The general proposi-

tion ‘When I reminisce, it is. . .’ in the English version

was retained in each of the items in the Chinese version.

Another research assistant, who was blind to the original

English version, then translated the scale back into

English. The Chinese version was carefully assessed by

the researchers to ensure its meaning was equivalent to

the English version. During the process, we removed one

of the items from the boredom reduction subscale (i.e.

When I reminisce, it is because it fills the gap when I find

time ‘heavy on my hands’.) in order to balance the num-

bers of items of the subscales. This was consistent with

recommendations made by Robitaille et al. (2010). The

resulting brief RFS consisted of 14 items with seven items

on the self-positive functions (i.e. problem solving: four

items; identity: three items) and other seven items on the

self-negative functions (boredom reduction: four items;

bitterness revival: three items).

The questionnaire was administered verbally in Can-

tonese. Trained interviewers conducted face-to-face

interviews to collect data by using standardized question-

naires, which assessed reminiscence, depressive symp-

toms, service utilization, health status, and demographic

characteristics. Variables related to reminiscence, depres-

sive symptoms, and demographic characteristics were

2 V.W.Q. Lou and J.C.P. Choy

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14

analyzed for the current study. Informed consent was

obtained from the participants before the administration

of the questionnaire.

Measures

Reminiscence functions. The brief 14-item RFS was used

to measure reminiscence functions (Robitaille et al.,

2010): boredom reduction (four items), bitterness revival

(three items), problem solving (four items), and identity

(three items). For each item, participants indicated the fre-

quency with which they used that specific function of

reminiscence, using a 6-point scale (1 ¼ never to 6 ¼ very

frequently). Scores on the bitterness revival and identity

subscales ranged from 3 to 18, whereas those on the prob-

lem solving and boredom reduction subscales ranged

from 4 to 24.

Depressive symptoms. We used the Chinese version of

the Geriatric Depression Scale – short form (GDS-15;

Lee, Chiu, & Kwong, 1994) to assess the elders’ depres-

sive symptoms. It consisted of 15 items, with each item

asking about participants’ experiences of depressive

symptoms in the week preceding the interview; higher

scores indicated higher levels of depression. The reliabil-

ity of the scale as indicated by Cronbach’s alpha was .81.

Demographic characteristics measured included gen-

der, age, and level of education.

Data analyses

Confirmatory factor analysis (CFA) was conducted to test

the factorial structure of the brief RFS. First, we com-

pared fit indices for three- and four-factor models. Next,

we tested model fit by comparing fit indices for a one-

factor model based on direct observation and for a one-

factor model based on the second-order of a four-factor

structure. Figure 1 illustrates the proposed second-order

factor structure of the brief RFS. Amos 18.0 was used to

complete the analyses, and model fit was evaluated by

examining important indices: the chi-square likelihood

ratio statistic, the comparative fit index (CFI), the good-

ness-of-fit index (GFI), the incremental fit index (IFI),

the Tucker–Lewis index (TLI), the root mean square

error of approximation (RMSEA), the standardized root

mean square residual (SRMR), and the expected cross-

validation index (ECVI).

We tested the psychometric properties of the Chinese

RFS by analyzing its internal consistency, reliability, cri-

terion-related validity, and descriptive statistics via SPSS

20. Regression analysis was also conducted to further

examine the relationship of the reminiscence functions

and depressive symptoms.

Results

Factor structure of the brief RFS: comparing a four-factor

solution and a three-factor solution. The fit indices from

the CFA of the tested first-order factor models are shown

in Table 1. Among the first-order factor models, a four-

factor model, with x2(71) ¼ 217.17, CFI ¼ .97, GFI ¼ .

96, IFI ¼ .97, TLI ¼ .97, RMSEA ¼ .055, SRMR ¼ .035,

and ECVI ¼ .42, demonstrated better fit than the three-

factor solution. Factor loadings from the four-factor

model ranged from .63 to .90. The four-factor model repli-

cated the expected brief RFS structure, with four factors

extracted from the original scale. Moreover, medium-

sized correlations were observed between factors

(r ¼ .42–.66; Table 2).

Factor structure of the brief RFS: comparing a first-

order solution and a second-order solution. The second-

order model, when compared to the one-factor model

based on direct observation, evidenced better model fit

Figure 1. Illustration of the proposed second-order factorial structure of the brief RFS. Notes: Bo ¼ boredom reduction; Bi ¼ bitternessrevival; PS ¼ problem solving; Id ¼ identity.

Aging & Mental Health 3

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14

(Table 1). Modifications in two pairs of error covariances

were made according to the observation of the modifica-

tion indices. Each pair of the items belonged to the same

factor, so that correlation of error covariance was allowed

(e.g. items 11 and 26 in boredom reduction, and items 1

and 17 in problem solving). The modified model demon-

strated a sound model fit for explaining the data, with

x2(71) ¼ 212.16, CFI ¼ .97, GFI ¼ . 96, IFI ¼ .97, TLI ¼.97, RMSEA ¼ .054, SRMR ¼ .049, ECVI ¼ .42, and

high and statistically significant factor loadings (p < .001).

The factor loadings of the items that loaded on to the first-

order factors ranged from 0.62 to 0.90, whereas the load-

ings of these factors on to the second-order factors ranged

from 0.62 to 0.94 (Table 3). As suggested by previous

research (Marsh, 1985), a hierarchical model is preferred

over a lower order model when comparable fit indices

between models are observed.

Psychometric properties of the brief Chinese version

of the RFS. Internal consistency, as indicated by

Cronbach’s alpha, ranged from .75 to .90 for the four fac-

tors (e.g. subscales) and 0.91 for the RFS. Correlations

between the four factors and depressive symptoms were

computed to examine criterion validity; the results are

listed in Table 2. Negative functions (boredom reduction

and bitterness revival) were moderately correlated with

depressive symptoms (r ¼ .33 and .35, respectively),

whereas positive functions (problem solving and identity)

demonstrated weak correlations (r ¼ .16 and .14, respec-

tively). The mean and standard deviation of the four sub-

scales and the RFS were also computed (Table 2).

Relationships of reminiscence functions with psycho-

logical well-being: multiple regressions on depressive

symptoms. Multiple regression analysis was conducted on

GDS with the four reminiscence functions as independent

variables. Results showed that negative functions

(boredom reduction and bitterness revival) contributed to

more depressive symptoms (b ¼ .24, p < .001; b ¼ .27,

p < .001). On the other hand, identity was found to be

negatively related to depressive symptoms (b ¼ �.12,

p < .05) (Table 4).

Discussion

The key objective of this study was to examine the facto-

rial structure of a brief RFS, consisting of the four sub-

scales (boredom reduction, bitterness revival, problem

solving, and identity), within a sample of older Chinese

adults. Results from a sample of 675 elders supported a

four-factor model solution. Furthermore, a second-order

model better represented the data, suggesting that the

four first-order factors loaded on to one second-order

factor.

Findings from the current study replicated the original

four-factor model (Robitaille et al., 2010; Webster, 1997),

indicating a sound cultural adaptation of the RFS scale for

the Chinese older adults. The problem solving and identity

subscales have been argued to be indistinct since their items

have loaded onto the same factor in exploratory factor anal-

yses of prior studies (Robitaille et al., 2010; Webster,

1993). Owing to this controversy, both three- and four-fac-

tor models were tested in the current study. We demon-

strated that a four-factor model was a better representation

of the data, suggesting that these two factors tended to be

distinct (as suggested by their theoretical meanings). Crite-

rion validity results were also consistent with previous

work (Cappeliez & O’Rourke, 2006). The results from

regression analysis showed that both negative functions

(boredom reduction and bitterness revival) have significant

positive association with depressive symptoms; whereas

identity demonstrates significant negative association with

depressive symptoms. This Chinese brief RFS can now be

Table 1. Summary of fit indices for the models tested.

Model Chi-square df CFI GFI IFI TLI RMSEA SRMR ECVI

Three-factor model 334.03��� 74 0.95 0.93 0.95 0.94 0.072 0.0467 0.59Four-factor model 217.17��� 71 0.97 0.96 0.97 0.97 0.055 0.0352 0.42

One-factor model based on direct observation

1760.19��� 77 0.69 0.67 0.69 0.63 0.18 0.1061 2.7

Second-order one-factor model 267.52��� 73 0.96 0.95 0.96 0.96 0.063 0.0516 0.49Second-order one-factor model with

modification212.16��� 71 0.97 0.96 0.97 0.97 0.054 0.0489 0.42

Note: ���p < .001.

Table 2. Correlation coefficients and internal consistency (N ¼ 675).

Cronbach’s alpha M (SD) 2 3 4 GDS

1. Reminiscence .91 22.9 (9.9)2. Boredom reduction .90 6.9 (3.9) .333. Bitterness revival .89 5.1 (3.0) .54 .354. Problem solving .82 6.0 (3.0) .48 .42 .165. Identity .75 4.9 (2.5) .55 .46 .66 .14

Note: All correlation coefficients were significant at the p < .001 level.

4 V.W.Q. Lou and J.C.P. Choy

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14

used with confidence to study the relationships between

functions of reminiscence and psychological well-being.

Moreover, findings of the present study supported a

second-order factor model of the brief RFS, thereby

expanding the measurement and conceptualization areas

of the reminiscence literature. Responses to the brief RFS

could be explained by four first-order factors (i.e. bore-

dom reduction, bitterness revival, problem solving, and

identity) and one second-order factor (i.e. reminiscence)

underlying these four factors. Thus, these distinct four-

factor factorial constructs are related and can be explained

by one common underlying higher level factor. When

compared to a first-order model with correlated factors, a

second-order model is more parsimonious and provides

theoretically error-free estimates of both the higher-order

general factor and each specific factor (Chen, Sousa, &

West, 2005). Theoretically, these results imply that remi-

niscence can function as an aggregated function on top of

four distinct functions, a result that should be considered

when designing future intervention programs.

A few limitations of the current study should be

noted. First, a loss of the original meaning in some items

might not have been fully avoided during the translation

of the RFS from English to Chinese: there might have

still been semantic or linguistic discrepancies between

the original and the targeted language (Bontempo,

1993). Second, all participants included in the current

study were elders living alone or only with their spouses.

Third, even though the four functions of reminiscence

included in the present study was selected purposefully,

there is the possibility that other prosocial reminiscence

functions (i.e. teach/inform and conversation) are also

associated with depressive symptoms. Manifestation of

these functions and their relationship with psychological

well-being in Chinese older adults deserve further study.

Therefore, interpretations and/or generalizations from

these data must be made with caution. Studies with

larger and more diverse samples are needed to replicate

our findings.

Conclusion

The current study is the first of its kind to examine remi-

niscence functions among Chinese older adults. Our find-

ings support a four-factor model of the brief RFS with

Chinese older adults, distinguishing the problem solving

and identity functions of reminiscence. More importantly,

a second-order model comprising one second-order factor

was established, extending previous literature on the con-

ceptualization and measurement of the functions of remi-

niscence. Finally, the reliability and validity of the brief

Chinese version of the RFS were satisfactory, supporting

its future use in research and practice among Chinese

older adults.

Table 4. Multiple regression on GDS.

95% CI for b

Factor b SE b Lower Upper

Reminiscence functionsBoredom reduction 0.22 0.04 0.24��� 0.14 0.30Bitterness revival 0.32 0.05 0.27��� 0.22 0.42Problem solving 0.01 0.06 0.01 �0.10 0.12Identity �0.17 0.07 �0.12� �0.31 �0.03

R2– 0.16���

Notes: �p < .05; ���p < .001.

Table 3. Factor loadings of the second-order model of the brief RFS.

RFS item Standardized Unstandardized (SE)

Boredom reduction 0.73 0.95 (0.078)11b. To pass the time during idle or restless hours 0.89 1.33 (0.055)19. To reduce boredom 0.84 1.12 (0.046)26b. For lack of any better mental simulation 0.88 1.13 (0.047)29a. For something to do 0.78 1.00

Bitterness revival 0.64 1.04 (0.089)7. To remember an earlier time when I was treated unfairly by others 0.76 0.79 (0.032)15. To rekindle bitter memories 0.89 1.02 (0.033)23a. To keep memories of old hurts fresh in my mind 0.90 1.00

Problem solving 0.84 0.76 (0.066)1b. To remind me that I have the skills to cope with present problems 0.69 1.28 (0.087)9. To put current problems in perspective 0.79 1.21 (0.073)17b. To see how my strengths can help me solve a current problem 0.74 1.22 (0.079)25a. To avoid repeating past mistakes at some later date 0.66 1.00

Identity 0.94 1.002. Because it helps me contrast the ways I have changed with the ways

I have stayed the same0.77 1.23 (0.082)

10. Because remembering my past, helps me define who I am now 0.76 1.07 (0.072)18a. As a means of self-exploration and growth 0.62 1.00

Notes: aParameters fixed to the value of 1.00. bItem pair with their error of covariance correlated.Source of item wordings: Robitaille et al. (2010).

Aging & Mental Health 5

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14

Acknowledgements

The authors would like to express their gratitude to Prof. Phil-ippe Cappeliez, who provided valuable comments on a priordraft of the manuscript.

Funding

This work was supported by the Hong Kong Jockey Club Chari-ties Trust, CADENZA Community Project.

References

Bluck, S., & Levine, L.J. (1998). Reminiscence as autobiograph-ical memory: A catalyst for reminiscence theory develop-ment. Ageing & Society, 18(2), 185–208.

Bontempo, R. (1993). Translation fidelity of psychologicalscales: An item response theory analysis of an individual-ism-collectivism scale. Journal of Cross-Cultural Psychol-ogy, 24(2), 149–166.

Cappeliez, P., & O’Rourke, N. (2002). Personality traits and exis-tential concerns as predictors of the functions of reminiscencein older adults. The Journals of Gerontology: Series B: Psy-chological Sciences and Social Sciences, 57(2), P116–P123.

Cappeliez, P., & O’Rourke, N. (2006). Empirical validation of amodel of reminiscence and health in later life. The Journalsof Gerontology: Series B: Psychological Sciences and SocialSciences, 61(4), P237–P244.

Cappeliez, P., O’Rourke, N., & Chaudhury, H. (2005). Functionsof reminiscence and mental health in later life. Aging & Men-tal Health, 9(4), 295–301. doi:10.1080/13607860500131427

Chen, F.F., Sousa, K.H., & West, S.G. (2005). Teacher’s corner:Testing measurement invariance of second-order factormodels. Structural Equation Modeling, 12(3), 471–492.

Coleman, P.G. (1986). Ageing and reminiscence processes:Social and clinical implications. Chichester: Wiley.

Conway, M.A., Wang, Q., Hanyu, K., & Haque, S. (2005). Across-cultural investigation of autobiographical memory onthe universality and cultural variation of the reminiscencebump. Journal of Cross-Cultural Psychology, 36(6), 739–749.

Cully, J.A., LaVoie, D., & Gfeller, J.D. (2001). Reminiscence,personality, and psychological functioning in older adults.The Gerontologist, 41(1), 89–95. doi:10.1093/geront/41.1.89

Goncalves, D.C., Guedes, J.M., Fonseca, A.M., & Martin, I.(2010). Psychometric properties of the Reminiscence Func-tions Scale for the Portuguese population: A preliminaryreport. International Journal of Aging & Human Develop-ment, 71(2), 153–166.

Korte, J., Bohlmeijer, E.T., Westerhof, G.J., & Pot, A.M. (2011).Reminiscence and adaptation to critical life events in olderadults with mild to moderate depressive symptoms. Aging &Mental Health, 15(5), 638–646.

Lee, H.-C.B., Chiu, H.F.K., & Kwong, P.P.K. (1994). Cross-val-idation of the Geriatric Depression Scale short form in theHong Kong elderly. Bulletin of the Hong Kong Psychologi-cal Society, 32, 72–77.

Lo Gerfo, M. (1980). Three ways of reminiscence in theory andpractice. The International Journal of Aging & Human Devel-opment, 12(1), 39–48. doi:10.2190/cbkg-xu3m-v3jj-la0y

Marsh, H.W. (1985). The structure of masculinity/femininity:An application of confirmatory factor analysis to higher-order factor structures and factorial invariance. MultivariateBehavioral Research, 20(4), 427–449.

Mezred, D., Petigenet, V., Fort, I., Blaison, C., & Gana, K. (2006).La r�eminiscence: Concept, fonctions et mesures. Adaptationfrancaise de la Reminiscence functions scale. [Concept, func-tions and measures of reminiscence: Adaptation of a Frenchversion of the Reminiscence Functions Scale]. Les CahiersInternationaux de Psychologie Sociale, 71, 3–14.

O’Rourke, N., Cappeliez, P., & Claxton, A. (2011). Functions ofreminiscence and the psychological well-being of young-oldand older adults over time. Aging & Mental Health, 15(2),272–281. doi:10.1080/13607861003713281

Robitaille, A., Cappeliez, P., Coulombe, D., & Webster, J.D.(2010). Factorial structure and psychometric properties ofthe reminiscence functions scale. Aging & Mental Health,14(2), 184–192. doi:10.1080/13607860903167820

Webster, J.D. (1993). Construction and validation of the Remi-niscence Functions Scale. Journal of Gerontology, 48(5),P256–P262.

Webster, J.D. (1997). The Reminiscence Functions Scale: A rep-lication. International Journal of Aging & Human Develop-ment, 44(2), 137–148.

Webster, J.D. (1998). Attachment styles, reminiscence function,and happiness in young and elderly adults. Journal ofAging Studies, 12(3), 315–330. doi:10.1016/s0890-4065(98)90006-8

Webster, J.D., & McCall, M.E. (1999). Reminiscence functionsacross adulthood: A replication and extension. Journal ofAdult Development, 6(1), 73–85.

Westerhof, G.J., Bohlmeijer, E., & Webster, J.D. (2010). Remi-niscence and mental health: A review of recent progress intheory, research and interventions. Ageing & Society, 30(4),697–721. doi:10.1017/s0144686X09990328

Wong, P.T., & Watt, L.M. (1991). What types of reminiscenceare associated with successful aging? Psychology and Aging,6(2), 272–279. doi:10.1037/0882-7974.6.2.272

6 V.W.Q. Lou and J.C.P. Choy

Dow

nloa

ded

by [

Viv

ian

W. Q

. Lou

] at

15:

30 0

7 Ja

nuar

y 20

14