HHS Public Access

Author manuscript
J Posit Psychol. Author manuscript; available in PMC 2021 December 08.
Author Manuscript

Published in final edited form as:

J Posit Psychol. 2021 ; 16(5): 651–659. doi:10.1080/17439760.2020.1791946.

Loved and lost or never loved at all? Lifelong marital histories

and their links with subjective well-being
Mariah F. Purol, Victor N. Keller, Jeewon Oh, William J. Chopik, Richard E. Lucas
Michigan State University

Abstract
Author Manuscript

Marriage has been linked to higher well-being. However, previous research has generally
examined marital status at one point in time or over a relatively short window of time. In order
to determine if different marital histories have unique impacts on well-being in later life, we
conducted a marital sequence analysis of 7,532 participants from the Panel Study of Income
Dynamics (54.2% women; Mage = 66.68, SD = 8.50; 68.7% White/Caucasian). Three different
marital sequence types emerged: a “consistently-married” group (79%), a “consistently-single”
group (8%), and a “varied histories” group (13%), in which individuals had moved in and out
of various relationships throughout life. The consistently-married group was slightly higher in
well-being at the end of life than the consistently-single and varied histories groups; the latter two
groups did not differ in their well-being. The results are discussed in the context of why marriage
is linked to well-being across the lifespan.
Author Manuscript

Keywords
sequence analysis; lifespan approach; marital history; subjective well-being; life satisfaction

The preponderance of evidence suggests that married people are happier than single people
(e.g., Perelli-Harris, Hoherz, Lappegard, & Evans, 2019). However, much of the existing
research on this topic only examines the marital status of an individual at one particular
time point and relates this single measurement to their happiness. Yet, people often move in
and out of relationships throughout their lives. It is not yet known whether these transitions
are associated with differences in well-being. For example, are people who experience
many marital transitions (multiple marriages and divorces) as happy as those who have
stayed single their whole lives? In the current study, we examined the marital histories of
Author Manuscript

7,532 individuals, followed from ages 18 to 60, and determine whether these histories were
differentially associated with life satisfaction in older adulthood.

The Link Between Marriage and Well-Being

Prior research consistently shows that married people report higher well-being than those
who are single, divorced, or widowed (Waite & Gallagher, 2001). Indeed, being married
is associated with several emotional, social, and health-related benefits across the lifespan

Correspondence concerning this manuscript should be addressed to Mariah F. Purol, Department of Psychology, Michigan State
University, 316 Physics Rd., East Lansing, Michigan 48824. purolmar@[Link].
 Purol et al. Page 2

(Haring-Hidore, Stock, Okun, & Witter, 1985). For example, married individuals report
Author Manuscript

less anxiety, depression, and mental distress (Ross, Mirowsky, & Goldstein, 2009), are
healthier and live longer (Schone & Weinick, 1998), and adjust well to the onset of chronic
illness (Revenson, 1994). Given these consistent associations, researchers have turned their
attention to identifying their causes.

There are many reasons why marriage might be associated with greater well-being.
Although the most obvious explanations point to the benefits of the companionship that
comes with marriage, alternative explanations for this association exist. Johnson and Wu
(2002), for instance, suggested that the association between marriage and subjective well-
being may actually reflect a different process—namely, that divorced individuals report
lower well-being than those who have not divorced, and they may be the ones primarily
driving the difference between groups of different marital status. Specifically, Johnson and
Wu proposed that the stress of the divorce and its consequences (e.g., social isolation, shared
Author Manuscript

childcare, economic hardship) can cause divorced individuals to experience more distress
than those who are married. Divorced individuals may also have poorer coping skills and
less social support, which likely amplifies the intensity of these stressors (Avison, 1999).
Altogether, the presence of these stressors likely lowers well-being.

Lucas and colleagues (2003, 2005) likewise suggested that these differences in happiness
may not reflect a causal relationship between marriage and happiness, but, instead, stable
individual differences. They argued that the association could be a result of selection effects:
those who end up getting married may have been happier than those who never do—even
before they got married. In examining how individuals adapt to changes in marital status,
they found evidence of this effect, such that happier people were more likely to enter into
and maintain a marriage than their less happy counterparts (who either did not marry or
Author Manuscript

eventually divorced). This suggests that there may be pre-existing differences in happiness
between those who remain married, singles, and those who eventually divorce.

It is also possible that the apparent effect of marriage is not due to marriage per se,
but to the existence of any close, interpersonal relationship (e.g., high quality friendships;
DePaulo & Morris, 2005). Singles and those in relationships can report similar levels of
well-being (Greitemeyer, 2009), perhaps because of positive relationships with their family
or friends (DePaulo & Morris, 2005). However, regardless of the reasons why, married
people generally report higher well-being than single, divorced, and widowed individuals in
cross-sectional comparisons.

Marital Transitions Across Life

Author Manuscript

Although the association between marital status and subjective well-being is robust, there
are still questions left unanswered. Importantly, many of the aforementioned studies merely
take a snap-shot of married and non-married people at one point in time. There have
been comparatively fewer longitudinal studies investigating how marital transitions impact
subjective well-being. Further, studies examining the entire marital history of an individual
are rare. Those studies that have been conducted suggest that the effects of marital status

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 3

on well-being can be quite complicated (e.g., Lucas, 2005; Lucas & Clark, 2006; Lucas &
Author Manuscript

Dyrenforth, 2005).

For example, divorce is typically perceived as a negative event, which might lead to the
expectation that divorced individuals will report lower subjective well-being than those who
are married. However, Lucas (2005) described levels of life satisfaction among participants
who were going through a divorce; this research showed that levels of life satisfaction
did not track changes in official marital status in a simple way. People who got divorced
experienced declines in life satisfaction even before the divorce officially occurred (i.e.,
when their marital status was still “married”). In addition, the divorce itself was associated
with an increase in well-being relative to lowest points at the end of the marriage, though
satisfaction levels were still lower than they were at the beginning of or prior to the
marriage. Further complicating these patterns, individuals who remarry experience a greater
recovery in life satisfaction after their divorce (albeit they do not quite return to their
Author Manuscript

baseline levels of life satisfaction).

Additionally, multiple marriage transitions impact life satisfaction in intricate ways.

Luhmann and Eid (2009) found that, for those who had undergone repeated divorces, life
satisfaction was higher at their second divorce than their first, and, for those who had
undergone more than one marriage, life satisfaction was similar across marriage events.
Taken as a whole, these results suggest that having an appreciation for the fact that marital
histories can be quite nuanced may provide additional insights into the association between
marital status and well-being.

Other observations suggest that the differences in well-being between those who are married
versus unmarried may not be as stark as some researchers have suggested (DePaulo,
Author Manuscript

2007). Specifically, research comparing married individuals and lifelong singles points to
a similarly complex pattern of associations between marital status and well-being. For
example, the association between marital status and well-being varies considerably across
contexts, including differences in the size of the association across countries (Lucas &
Dyrenforth, 2005). There is also evidence to suggest that the effects of marital status
on well-being might even be decreasing in recent cohorts, as people report being more
satisfied with being single (Boger & Huxhold, 2018). In sum, although cross-sectional
studies suggest a robust association between marital status and subjective well-being, the
nuance of the association requires a more sophisticated design. Examining an individual’s
entire marital history gives a more holistic picture of the various transitions they go through
over time.

The body of longitudinal research on marital transitions suggests an intricate relationship

Author Manuscript

with subjective well-being that has not yet been effectively captured. In order to precisely
determine how relationships impact happiness, longitudinal research needs to account for
the many marital transitions across the lifespan with individuals’ entire marital history.
Marital histories provide a year-by-year account of an individual’s relationship status,
including transitional periods. By getting a more holistic view of how individuals move
through relationships throughout life, we are better able to draw conclusions about these
relationships’ impact on well-being and happiness.

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 4

The Current Study

Author Manuscript

Although the relationship between marital status and life satisfaction has been widely
studied, past research has relied on a limited scope of individuals’ relationship histories,
failing to account for all of the relationship transitions a person may have undergone in
a lifetime. At the same time, it is possible that groups of individuals can be identified
based on whether or not they experience similar transitions across their lifespans. The
current study examines differences in life satisfaction among those with different marital
histories. In order to do this, we employed sequence analysis to thematically and empirically
group individuals who experienced similar marital transitions. Doing so provides valuable
descriptive data for how common different relationship transition patterns are. For example,
there could be a relatively homogenous group of people who have been married to one
person most of their adult life. Others could have experienced varied marital histories
consisting of multiple marriages and periods of being single. In a descriptive approach, the
Author Manuscript

sequence analysis allowed us to compare the marital histories of 7,532 participants followed
from ages 18 to 60 to determine which patterns were most common. We also examined
how people in these sequences differed from each other in well-being in late life in order
to provide a more comprehensive understanding of how marital status is associated with life
satisfaction.

Method
Participants
Participants were 7,532 respondents (54.2% women; Mage = 66.68, SD = 8.50) from the
Panel Study of Income Dynamics (PSID). The PSID is a longitudinal study that has
followed a nationally representative sample of U.S. individuals and their families since
Author Manuscript

1968. Children of original study participants (and their children) are subsequently included
in data collection as the survey progresses. Participants were 68.7% White/Caucasian, 28.9%
Black/African American, 1.7% Spanish American/Puerto Rican/Mexican/Cuban, and .7%
Other. On average, participants had 13.17 years (SD = 2.82) of education.

We examined marital histories for participants from ages 18 to 60, collected both
prospectively and retrospectively. Well-being information was available for a subset of the
participants (N = 3,530; 54% women; Mage = 66.23, SD = 8.01), as well-being was not
assessed until 2009. Participants in this subsample were 75.2% White/Caucasian, 22.9%
Black/African American, 1.4% Spanish American/Puerto Rican/Mexican/Cuban, and .6%
Other. Average level of education for the subsample was 13.26 years (SD = 2.80).
Author Manuscript

Measures
Martial histories.—Assessment of marital history data in the PSID began in 1985.
Participants in the 1985 wave were asked to give a retrospective account of their entire
marital history. In all successive waves, participants provided annual marital history updates,
providing information for their current/most recent marriage in regard to the previous
calendar year. Data and transitions after age 60 were sparse, so this age was used as a
cut-off to maximize our sample size.

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 5

Life satisfaction1.—Life satisfaction was measured with a single item, “Please think
Author Manuscript

about your life as a whole. How satisfied are you with it?” Responses ranged from 1
(completely satisfied) to 5 (not at all satisfied). The scale was reverse-coded so that higher
levels indicated greater life satisfaction. Single-item measures of life satisfaction have
comparable validity to longer form measures (Cheung & Lucas, 2014; Lucas & Donnellan,
2012). Life satisfaction was assessed at five time points (i.e., 2009, 2011, 2013, 2015,
2017). Because we were interested in participants’ life satisfaction at the end of the marital
sequence, we used the life satisfaction assessment most recently available after age 60. Thus,
participants in all analyses had full marital history from age 18 to age 60.

Results
Sequence Analysis
Individual marriage histories were organized into sequences of years. Each year, a person
Author Manuscript

could be in one of 6 states: never married (single), married, divorced, separated, widowed,
or deceased. All sequences started at age 18 and ended at age 60. To determine whether
there are common patterns of marital histories, we performed two analyses: one, a sequence
analysis, which assessed the similarities between the marriage histories, and, two, a cluster
analysis, to determine which clusters of marriage histories best summarize the data.

The sequences were then compared to each other using optimal matching (Abbott &
Forrest, 1986). This procedure determines the similarity of two sequences by counting the
fewest amount of insertions, deletions, and substitutions that are necessary to transform
one sequence into the other.2 For example, a four-year sequence of one individual (i.e.,
married married married married) might require one substitution to be transformed into the
sequence of another individual (i.e., married married married divorced). The more insertions,
Author Manuscript

deletions, or substitutions necessary to transform one sequence into another, the more
dissimilar they are. The optimal matching procedure was performed on all possible pairs of
sequences resulting in a matrix with a row and column for every sequence and cells with the
dissimilarity scores for any two sequences.

The cluster analysis recommended by Studer (2013) was then performed on the
dissimilarities to determine how the sequences could be best clustered. This analysis,
Partitioning Around Medoids (PAM) initialized with Hierarchical Clustering (HC), is
explained in detail by Studer (2013). The hierarchical clustering used the “Ward” algorithm.
Essentially, this analysis groups similar sequences into a predetermined number of clusters,
then optimizes the clusters by swapping sequences until the overall dissimilarity between
the sequences within a cluster is as low as possible. PAM initialized with HC was done for
Author Manuscript

solutions of 2 to 20 clusters.

To determine which number of clusters best represents the data, a set of clustering-quality
measures recommended by Studer (2013) were calculated. The measures used were the

1Worth noting, it wasn’t until 2009 that PSID started assessing life satisfaction, so plenty of participants were much older than age 60
(i.e., those born prior to 1949). As a result, the age at which this first assessment took place varies considerably. To account for this
variability, we controlled for age at assessment in all analyses.
2The costs for insertions, deletions, and substitutions were all set to 1.

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 6

Point Biserial Correlation (PBC), Hubert’s Gamma (HG), Hubert’s C (HC), and the Average
Author Manuscript

Silhouette Width (ASW). The PBC and HG have a similar interpretation: they are indices
of how well the clustering can reproduce the dissimilarity matrix. If a clustering solution
reproduces the dissimilarity matrix well, then it is considered to be a good clustering
solution. HC is a measure of the difference between the solution obtained and the best
solution theoretically possible (for this measure, lower numbers indicate higher clustering
quality). Finally, ASW indicates whether the within-cluster similarity is greater than the
between-cluster similarity.

The measures of clustering quality were highly correlated, and, as can be seen in Figure
1, they indicated that 3 clusters were the optimal solution for this data set of marriage
sequences. As seen in Figure 2, the resulting clusters were (1) people who spent most of
their adult lives married (“consistently-married;” 79.30% of the sample), (2) people who
spent most of their adult lives single (“consistently-single;” 7.86% of the sample), and (3)
Author Manuscript

people who transitioned throughout various relationship states (“varied-histories;” 12.84%

of the sample).

The individuals in these groups were not homogenous in their relationship histories. It is
important to note that cluster membership is assigned by the length of time spent in each
state, and not an individual’s exclusive association with this state. For example, while those
in the consistently-married cluster are characterized by their long marriages, the group
includes those who have been widowed or divorced as well. Similarly, those in consistently-
single cluster are characterized by long periods of singlehood, although some eventually
married, divorced, or became widowed.

Figure 3 shows the relative distribution of marital states for each cluster from ages 18 to
Author Manuscript

60 (i.e., chronograms). Consistent with our interpretation of the sequence graphs (Figure 2),
over time, a large proportion of people became married (in Cluster 1), a large proportion of
people stayed single (in Cluster 2), and the proportion of single and married individuals was
larger earlier in life, followed by a larger proportion of divorces (in Cluster 3).

Linear Regression
We then compared the life satisfaction of individuals from the different clusters. To do
so in the context of linear regression, one must pick a reference group against which to
compare group differences in life satisfaction (i.e., a dummy coded approach). We chose
the consistently-married group because they had the largest number of participants available.
Thus, the effects of consistently-single and varied-histories groups were estimates of the
differences between these groups and the consistently-married group. We controlled for
Author Manuscript

gender and education, as these factors are known to predict life satisfaction in different
ways (Chipperfield & Havens, 2001; Fernández-Ballesteros, Zamarrón, & Ruiz, 2001).
Specifically, men are more likely to get remarried than women after a divorce (Bradbury &
Karney, 2010), and those with lower levels of education often have shorter-lasting marriages
(Wang, 2015). Thus, we included gender and education as control variables.

The results of this analysis are presented in Table 1. Compared to the consistently-married
group, individuals in the consistently-single group and varied-histories group reported lower

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 7

life satisfaction. People with higher levels of education also reported higher life satisfaction,
Author Manuscript

which is consistent with prior research (Melin, Fugl-Meyer, & Fugl-Meyer, 2003). In a
follow-up analysis, we varied the reference group such that the comparison was against
the varied-histories group (see Table 2). Reproducing the effect above, people in the
consistently-married group reported higher levels of happiness relative to the varied-history
group. Importantly, the consistently-single and varied-history groups did not significantly
differ in their levels of life satisfaction.

Individuals in the consistently-married group reported higher level of life satisfaction than
individuals either of the other groups. To contextualize these effects, the differences were
relatively small in magnitude (βs > .05). This suggests that there are not dramatic differences
between people with dramatically different marital histories.

Discussion
Author Manuscript

The current study used sequence analysis to examine differences in life satisfaction of
individuals with different marital histories. This method allowed us to determine common
marital transition patterns and get a more holistic picture of the association between
relationship history and life satisfaction. We found that individuals generally fall into one
of three marital history groups: consistently-married (married for most of their adult life),
consistently-single (single for most of their adult life), or varied-histories (demonstrated
varied patterns of relationships throughout their adult life). The large majority of participants
fell into the consistently-married group. When examining life satisfaction, we found that the
consistently-married individuals had higher life satisfaction compared to consistently-single
and varied-history individuals, but the consistently-single individuals and varied-histories
groups did not differ in their life satisfaction.
Author Manuscript

The Link between Marital Histories and Well-being

Our analyses showed that, when using a life history approach, greater time spent married
was positively associated with well-being towards the end of life. Those who had spent
much of their adult lives single and those who experienced many transitions were similar in
their levels of well-being. These findings are consistent with previous research in two ways.
First, they are consistent in direction: marriage was associated with higher life satisfaction
compared to divorce/singlehood. This is consistent with a large body of literature showing
higher levels of life satisfaction among married individuals. Second, these associations are
consistent in magnitude: the differences in life satisfaction between these groups, although
significant, were relatively small. This, too, is consistent with previous research (Boger &
Huxhold, 2019; Lucas & Dyrenforth, 2005).
Author Manuscript

While the reports of greater life satisfaction of married individuals are relatively
unsurprising, the lack of difference between the consistently-single and varied-histories
groups was unanticipated. As previously discussed, those who remain married and those
who eventually get divorced have been shown to differ in life satisfaction before the
marriage event (Lucas, 2005). The similarity in life satisfaction in older adulthood between
the consistently-single and varied-history groups could likewise be attributable to similarity
in life satisfaction earlier in life of singles and individuals who eventually divorce.

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 8

Other possibilities exist as well, although, at the moment, they are a bit speculative. For
Author Manuscript

example, individuals who eventually get divorced may indeed be happier than single people
earlier in life (e.g., before or when they are married). However, given the persistent declines
in well-being that accompany divorce, consistently-single and varied-history individuals
may end up looking relatively similar with respect to well-being. The negative feelings
that can accompany divorce, such as feelings of failure, emotional instability, depression,
or guilt (Thomas & Ryan, 2008) may prompt an enduring decrease in well-being. In other
words, decreased well-being may be a lasting response to divorce, which causes divorced
individuals’ well-being to match that of singles.

Yet another explanation for this similarity is that consistently–single and varied-history
individuals may both struggle with social isolation, albeit from different sources. Spielmann
and colleagues (2013) posited that some single people may experience a “fear or being
single”, fueled, in part, by a sense of loneliness or apprehension about being “alone forever.”
Author Manuscript

These authors also suggested that those who are very fearful of being single may “settle”
in a relationship—sometimes choosing partners that are less responsive or attractive, or
pine for previously unsuccessful or problematic relationships from their past (Spielmann,
MacDonald, Joel, & Impett, 2016; Spielmann et al., 2013). Social isolation might also be
salient for those who experience divorce (Johnson & Wu, 2002). In addition to losing a
spouse, those who are divorced may also lose ties with family or adult children (de Jong
Gierveld, van Tilburg, & Dykstra, 2006). Loneliness caused by this loss of connections may
worsen with age, as well (de Jong Gierveld, van Tilburg, & Dykstra, 2006). For both of these
groups (singles and those divorced), social isolation or loneliness maybe a common factor in
decreasing well-being. However, future research is needed in order to definitively determine
the sources of life satisfaction within these separate groups.
Author Manuscript

Limitations and Future Directions

Our study examined complete marital histories from a large sample of individuals, it used
an innovative data reduction technique (e.g., sequence analysis) to make sense of marital
transitions, and it linked variation in these transitions to well-being in older adulthood. Yet,
despite these strengths, some limitations do exist. Unlike previous studies (Lucas, 2005;
Luhmann & Eid, 2009), we did not have measures of life satisfaction before and after
each marital transition. Rather, our approach allowed us to examine the culmination of
all marriage events across the entirety of a person’s life. We viewed these approaches as
related but not entirely overlapping. One approach allows researchers to examine short-term
pre- and post-marital changes in life satisfaction in a relatively isolated time window. Our
approach allowed us to characterize the marital transitions of a lifespan and examined
whether or not people were happy at the end of their histories. Our approach leaves
Author Manuscript

open the possibility that the differences observed later in life might at least partially
reflect differences that may be observed earlier in life. Nevertheless, including more dense
measures of life satisfaction would have allowed us to parse apart the impact of each event
and re-occurrences of events (e.g., Luhmann & Eid, 2009).

Another limitation is that we examined an isolated life event (e.g., marital status) at each
age from 18 to 60. However, these changes in marital status occur in the context of other

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 9

life events that may have an impact of life satisfaction (e.g., parenthood, unemployment,
Author Manuscript

disability). Indeed, many of these life events have implications for marriage as well.
Unemployment causes a great deal of stress which may have negative effects on an
individual’s marriage (Kraft, 2001; Lester, 1996). Married, divorced, and single life all differ
in important ways depending on whether these people are also caring for children (Gringlas
& Weinraub, 1995; Rankin & Maneker,1985). Thus, the patterns of changes in marital
status that we observed may result from other important events—events that are causally
linked with the different levels of well-being that we identified across groups. Fortunately,
there are a number of ways in which sequence analysis can be adapted to account for the
life transitions of different events that occur simultaneously (e.g., multichannel sequence
analysis; Gauthier, Widmer, Bucher, & Notredame, 2010; Roux, Grimaud, & Leray, 2018).
Future research should examine whether there are additional descriptive patterns of life
transitions when other, related lifespan transitions are considered.
Author Manuscript

Conclusion
The current research found that, in general, people fall into one of three types of marital
histories: consistently-married, consistently-single, or varied-histories. Life satisfaction at
end of life differed depending on membership in these groups, with consistently-married
individuals being slightly happier than consistently-single or varied-histories individuals,
who did not differ from each other in life satisfaction. There are a variety of explanations
for these results: marriage may indeed lead to higher life satisfaction, those who are
consistently-single or have varied histories might experience similar social isolation or
additional stress, or these groups may differ from each other before these transitions even
begin. Additional research is needed in order to determine precisely why these groups are
similar or different, the implications of these differences, and will help further elucidate the
Author Manuscript

relationship between marital history and life satisfaction.

Acknowledgments
This study was not preregistered.

Research reported in this publication was supported by the National Institute of Aging of the National Institute of
Health under Award Numbers P01AG029409, R01-AG051142, and 2R03AG054705-01A1. The content is solely
the responsibility of the authors and does not necessarily represent the official views of the National Institutes of
Health. Portions of this research were presented at the 2020 SPSP (New Orleans, LA) Conference.

References
Abbott A, & Forrest J (1986). Optimal matching methods for historical sequences. The Journal of
Interdisciplinary History, 16(3), 471–494.
Author Manuscript

Avison WR (1999). Family structure and processes. In Horwitz AV & Scheid TL (Eds.), A handbook
for the study of mental health: Social contexts, theories, and systems (pp. 228–240). New York, NY:
Cambridge University Press.
Boger A, & Huxhold O (2018). The changing relationship between partnership status and loneliness:
effects related to aging and historical time. The Journals of Gerontology: Series B: Psychological
Sciences and Social Sciences. doi:10.1093/geronb/gby153
Bradbury TN & Karney BR (2010). Intimate relationships. New York, NY. WW. Norton and Company.
Cheung F, & Lucas RE (2014). Assessing the validity of single-item life satisfaction measures: results
from three large samples. Quality of Life Research, 23(10), 2809–2818. [PubMed: 24890827]

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 10

Chipperfield JG, & Havens B (2001). Gender differences in the relationship between marital status
transitions and life satisfaction in later life. The Journals of Gerontology Series B: Psychological
Author Manuscript

Sciences and Social Sciences, 56(3), 176–186.

de Jong Gierveld J, van Tilburg TG, & Dykstra PA (2006). Loneliness and social isolation. In
Perlman D & Vangelisti A (Eds.), The Cambridge handbook of personal relationships (pp. 485–
500). Cambridge, UK: Cambridge University Press.
DePaulo BM (2007). Singled out: How singles are stereotyped, stigmatized, and ignored, and still live
happily ever after. New York, NY: St. Martin’s Press.
DePaulo BM, & Morris WL (2005). Singles in society and in science. Psychological Inquiry, 16(2–3),
57–83.
Fernández-Ballesteros R, Zamarrón MD, & Ruiz MÁ (2001). The contribution of socio-demographic
and psychosocial factors to life satisfaction. Ageing & Society, 21(1), 25–43.
Gauthier J-A, Widmer ED, Bucher P, & Notredame C (2010). Multichannel Sequence Analysis
Applied to Social Science Data. Sociological Methodology, 40(1), 1–38.
Greitemeyer T (2009). Stereotypes of singles: Are singles what we think? European Journal of Social
Psychology, 39(3), 368–383.
Author Manuscript

Gringlas M, & Weinraub M (1995). The more things change...single parenting revisited. Journal of
Family Issues, 16(1), 29–52.
Haring-Hidore M, Stock WA, Okun MA, & Witter RA (1985). Marital status and subjective well-
being: A research synthesis. Journal of Marriage and the Family, 47(4), 947–953.
Johnson DR, & Wu J (2002). An empirical test of crisis, social selection, and role explanations of the
relationship between marital disruption and psychological distress: A pooled time‐series analysis
of four‐wave panel data. Journal of Marriage and Family, 64(1), 211–224.
Kraft K (2001). Unemployment and the separation of married couples. Kyklos, 54(1), 67–88.
Lester D (1996). The impact of unemployment on marriage and divorce. Journal of Divorce &
Remarriage, 25(3–4), 151–154. [PubMed: 12179707]
Lucas RE (2005). Time does not heal all wounds: A longitudinal study of reaction and adaptation to
divorce. Psychological Science, 16(12), 945–950. [PubMed: 16313658]
Lucas RE, & Clark AE (2006). Do people really adapt to marriage? Journal of Happiness Studies, 7(4),
405–426.
Author Manuscript

Lucas RE, & Donnellan MB (2012). Estimating the reliability of single-item life satisfaction measures:
Results from four national panel studies. Social Indicators Research, 105(3), 323–331. [PubMed:
23087538]
Luhmann M, & Eid M (2009). Does it really feel the same? Changes in life satisfaction following
repeated life events. Journal of Personality and Social Psychology, 97(2), 363. [PubMed:
19634981]
Luhmann M, Hofmann W, Eid M, & Lucas RE (2012). Subjective well-being and adaptation to life
events: a meta-analysis. Journal of Personality and Social Psychology, 102(3), 592. [PubMed:
22059843]
Melin R, Fugl-Meyer KS, & Fugl-Meyer AR (2003). Life satisfaction in 18-to 64-year-old Swedes: in
relation to education, employment situation, health and physical activity. Journal of Rehabilitation
Medicine, 35(2), 84–90. [PubMed: 12691338]
Perelli-Harris B, Hoherz S, Lappegård T, & Evans A (2019). Mind the “happiness” gap: the
relationship between cohabitation, marriage, and subjective well-being in the United Kingdom,
Australia, Germany, and Norway. Demography, 56(4), 1219–1246. [PubMed: 31290087]
Author Manuscript

Revenson TA (1994). Social support and marital coping with chronic illness. Annals of Behavioral
Medicine, 16(2), 122–130.
Rankin RP, & Maneker JS (1985). The duration of marriage in a divorcing population: The impact of
children. Journal of Marriage and the Family, 47(1) 43–52.
Ross CE, Mirowsky J, & Goldsteen K (1990). The impact of the family on health: The decade in
review. Journal of Marriage and the Family, 52(4), 1059–1078.

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 11

Roux J, Grimaud O, & Leray E (2018). Multichannel sequence analysis: An innovative method
to study patterns of care pathways. Application to multiple sclerosis based on French Health
Author Manuscript

Insurance data. Revue d'Épidémiologie et de Santé Publique, 66, 430–431.

Schone BS, & Weinick RM (1998). Health-related behaviors and the benefits of marriage for elderly
persons. The Gerontologist, 38(5), 618–627. [PubMed: 9803650]
Spielmann SS, MacDonald G, Maxwell JA, Joel S, Peragine D, Muise A, & Impett EA (2013). Settling
for less out of fear of being single. Journal of Personality and Social Psychology, 105(6), 1049.
[PubMed: 24128187]
Studer M (2013). WeightedCluster library manual: A practical guide to creating typologies
of trajectories in the social sciences with R. Retrieved from [Link]
unige:78576
Waite LJ, & Gallagher M (2001). The case for marriage: Why married people are happier, healthier,
and better off financially. New York, NY: Random House Digital, Inc.
Wang W (2015). The link between a college education and a lasting marriage. Washington, DC: Pew
Research Center.
Author Manuscript
Author Manuscript
Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 12
Author Manuscript
Author Manuscript

Figure 1.
Clustering quality for 2 to 20 clusters
Note. Quality indicators are standardized.
Author Manuscript
Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 13
Author Manuscript
Author Manuscript

Figure 2.
Author Manuscript

Index plots of the three clusters of marital histories

Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 14
Author Manuscript
Author Manuscript

Figure 3.
Author Manuscript

Chronograms (relative frequencies of states) of the three clusters of marital histories

Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 15

Table 1.

Regression predicting subjective well-being from cluster membership

Author Manuscript

b SE β t p LB UB
Intercept 3.807 0.158 24.106 < .001 3.497 4.117
Consistently-Single −0.184 0.058 −0.054 −3.153 .002 −0.299 −0.070
Varied History −0.301 0.043 −0.120 −7.010 < .001 −0.385 −0.217
Age −0.001 0.002 −0.006 −0.368 .713 −0.004 0.003
Gender −0.018 0.028 −0.011 −0.628 .530 −0.074 0.038
Education 0.016 0.005 0.053 3.080 .002 0.006 0.026

Note. Consistently-married cluster served as the reference group. Gender: 1=men, 2=women. LB: Lower bound of 95% confidence interval; UB:
Upper bound of 95% confidence interval.
Author Manuscript
Author Manuscript
Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.

 Purol et al. Page 16

Table 2.

Regression predicting subjective well-being from cluster membership

Author Manuscript

b SE β t p LB UB
Intercept 3.506 0.160 21.960 < .001 3.193 3.819
Consistently-single 0.117 0.069 0.034 1.704 .088 −0.018 0.251
Consistently-married 0.301 0.043 0.141 7.010 < .001 0.217 0.385
Age −0.001 0.002 −0.006 −0.368 .713 −0.004 0.003
Gender −0.018 0.028 −0.011 −0.628 .530 −0.074 0.038
Education 0.016 0.005 0.053 3.080 .002 0.006 0.026

Note. Varied History cluster served as the reference group. Gender: 1=men, 2=women. LB: Lower bound of 95% confidence interval; UB: Upper
bound of 95% confidence interval.
Author Manuscript
Author Manuscript
Author Manuscript

J Posit Psychol. Author manuscript; available in PMC 2021 December 08.