The present study expands on the findings of Seligman, Steen, Park and Peterson (2005) who noted significant increases in participant’s happiness after recalling three pleasant events from their day in comparison to those who recalled events from childhood for the duration of a week. The goal of this study was to support these findings and also to determine whether sense of coherence (SOC) and frequency of writing responses mediated this increase in happiness.
Methodology: Sixty seven participants (30 male, 37 female) were randomised to write either three good things in their day (experimental group) or pleasant childhood memories (control condition) for the duration of fourteen days.
Results: The research found no significant increase in happiness or SOC in those who recalled three good things in their day. Thus, the role of SOC and frequency of writing in regard to mediating an increase in well-being was not established.
Conclusions: The study experienced a high rate of attrition where 41 participants dropped out (the final sample being 26). Future studies should increase their sample size so to improve the study’s statistical power and lessen the effects of attrition.
How do people adapt their lifestyle to enjoy a more enhanced and happier life? More specifically, what intrinsic factors allow for this improvement in happiness? It is thought that many people are in search of ‘flow’ so that they can be challenged while remaining competent in daily tasks and events (Diener, Lucas & Shigehiro, 2002). Happiness has been defined as an “overall satisfaction with our lives or as high average levels of enjoyment and other desired emotions” (Martin, p. 92, 2007). The discipline of positive psychology aims to unite both theory and research so to understand what makes life most meaningful, so to allow individuals and communities to flourish (Seligman, Steen, Park and Peterson, 2005). Positive psychology opposes the pathogenic approach to understanding human behaviour which underlies much psychological research by focusing on factors that contribute to and prevent mental illness. In contrast, it appropriates a salutogenic approach, researching how people maintain and enhance their well-being. These findings are easily applied to a real life context, particularly through the development of positive psychology interventions (PPI’s) (Seligman 2002). This research will discuss literature that supports the use of PPI’s, specifically examining the work of Seligman, Steen, Park and Peterson (2005). The present study is an expansion of their findings. Factors that mediate an increase in well-being will examined, where sense of coherence is thought to be the most appropriate (Antonovsky, 1987). The terms happiness and well-being will be used interchangeably as both are thought to measure the same construct. Furthermore, studies have found that using explicit measures of happiness are a good predictor of well-being (Kahneman & Krueger, 2006).
PPI’s enhance people’s well-being and happiness by developing existing virtues and strengths (Synder & Lopez, 2005). The aim of a PPI is to change your thought patterns in an effort to increase levels of well-being. A number of studies have successfully improved people’s well-being; through enhancing levels of optimism and gratitude (King 2001, Lyubomirsky et al., 2006), or by asking people to carry out acts of kindness (Dunn, Aknin & Norton 2008). Sin and Lyumborisky (2009) recently conducted a meta-analysis study investigating the effectiveness of PPI’s. Findings verified that fifty-one studies significantly increased well-being over a sustainable period of time. These studies indicate that PPI’s can provide a cost effective method of increasing well-being and can be applied in a number of settings, for example in community and clinical vicinities. Furthermore, they can help people resolve feelings of worthlessness before these feelings escalate into a significant problem. This style of intervention is also suitable for people who are affected by mental illness as it reduces the likelihood of post-treatment relapse (Snyder and Lopez, 2005). Despite significant advancements made in regard to PPI’s, research has yet to determine what intrinsic factors are responsible for this increase in well-being.
The current study is an expansion of research conducted by Seligman, Steen, Park and Peterson (2005). In their study, participants were asked to write about either three good things in their day and the reasons for these (experimental group) or events from their childhood (control group) for the duration of a week. Participants who wrote about three good things in their day reported significant increases in happiness and decreases in depression when compared to the control group. These changes were most evident one month following the intervention.
Following a thorough analysis of positive psychology literature, it is apparent that there is a disparity of research investigating what psychological mechanisms underlie these improvements in well-being, following a PPI. Seligman et al., (2005) argue that consistently identifying good things in one’s day leads to a greater sense of meaning in daily life. By determining what mediates this increase in well-being, greater validation and generalisability of this PPI would be feasible, contributing to the development and application of positive psychology in a broader context.
A possible explanation for this increase in well-being is the broaden and build theory which was developed by Frederickson (2004). The broaden and build theory suggests that people who experience high levels of positive affect undergo broadened thought-action repertoires. These increases in positive affect improve one’s general coping resources, allowing for better management of stressful situations (Frederickson, 2004). Yet, this theory is somewhat limited in explaining the current phenomenon as it is based on understanding how one can avoid illness by means of personal resilience which opposes the salutogenic approach of positive psychology. In addition, Frederickson’s theory is based on understanding the effects of positive affect. This excludes other factors that may be responsible for this increase in well-being, for example, increased feelings of life meaning, changes in coping mechanisms and thought processes. For these reasons, the broaden and build theory cannot be considered applicable to explaining what mediates this increase in well-being.
Alternatively, Lazurus and Folkman’s (1984) coping theory hypothesise that one may perceive their life situation more positively through reappraisal of negative situations. The coping theory focuses on three main factors; how one appraises a situation, the availability of coping resources such as social support and the coping mechanisms employed. Folkman et al., (1991) developed an intervention which attempted to change participant’s appraisal of situations from negative to positive and encouraged participants to adopt appropriate methods of coping. This intervention effectively increased feelings of positive morale and decreased symptoms of depression. In addition, Lyubomirsky et al., (2005) noted that forty per cent of individual differences in happiness are an outcome of one’s activities and perceptions of life circumstances (Snyder and Lopez, 2002). This approach could provide a valid explanation for an increase in well-being as one may appraise daily life more positively after recalling three good things that happened to them. However, the coping theory is also derived from a pathogenic perspective. Through investigating similar theoretical perspectives, Antonovsky’s (1987) sense of coherence is seen to provide a credible explanation for this increase in well-being for a number of different reasons.
Sense of coherence (SOC) supports the fundamental basis of positive psychology as it originates from a salutogenic approach and is concentrated on understanding how people manage and overcome stressful situations. The theory of SOC was developed by Aaron Antonovsky (1987) who wanted to understand why some people became ill, while others maintained their health when exposed to similarly stressful environments (Drageset, Nyaard, Erick, Bondevik, Nortvedt & Natvig, 2008). He recognised that health is based on an ease-disease continuum and his research has focused on the maintenance and enhancement of health (Holmberg, Thelin & Striernstrom, 2004). SOC is considered “a resource that enables people to manage tension, to reflect about their external and internal resources, to identify and mobilise them, to promote effective coping by finding solutions and resolve tension in a health promoting manner” (Eriksson and Lindstrom, p. 6 2006). It is important to recognise that SOC was not devised to predict behaviour but rather to provide an explanation of how people perceive life situations (Geyer, 1997). Moreover, it does not relate to a specific coping strategy but provides a basis for effective coping mechanisms (Gana, 2000).
This theory is comprised of three underlying constructs; comprehensibility, manageability and meaningfulness. These constructs contribute so to provide consistent life experiences where one can interpret their life accordingly. Comprehensibility is also referred to as cognitive controllability, scoring high in this construct increases the likelihood of perceiving one’s internal and external environment as predictable, understandable and expected (Geyer, 1997). SOC accounts for the likelihood of an unexpected event and subsequently developed the construct manageability. Manageability is a behaviour orientated construct where it is based on the premises that one perceives that he/she has the resources available so to deal with stressful circumstances. Meaningfulness is the motivational component of SOC where one appraises stressful situations as challenging and worthy of investment and commitment (Cohen, Ben Zur & Rosenfeld, 2008; Linstrom & Eriksson 2006).
Antonovsky’s theory also establishes how one’s level of SOC is maintained through the use of generalised resistance resources (GRR’s). GRR’s are an additional motivational element of SOC and allow for the advancement of health and well-being. They encompass any factor that lead to effective coping so that one gains an impression of “the world as ‘making sense’, cognitively, instrumentally and emotionally” (Antonovsky, p. 15, 1996). GRR’s comprise of both internal and external factors, including material factors (education, income); biological factors (intelligence, personality) and psychosocial factors (marital status, spirituality) (Lindstrom & Eriksson, 2006). Their purpose is to support a strong SOC by moderating and protecting one’s well-being so that one can deal effectively with demanding situations (Menzies, 2000). The availability of GRR’s when in a stressful situation is important as repeated absence of these resources may weaken one’s SOC (Johnson, Segesten & Mattsson, 2002). Overall, through examining the purpose of GRR’s, it could be theorised that identifying three good things daily functions like a GRR as one is reflecting on the positive aspects of their life. This leads to the perception that his/her life is comprehensive, manageable and meaningful.
To add to this explanation of increased well-being, numerous studies have indicated that levels of SOC are liable to change following significant life events. In the past, Antonovsky (1987) argued that SOC becomes stable at the age of thirty, yet this is considered vague and has been refuted by many (Geyer, 1997). For instance, in a four year longitudinal study Smith, Breslin and Beaton (2003) found that over a third of the sample reported a change in their SOC. Studies have indicated that a change in SOC is likely to occur following a traumatic incident as the event is seen to impact one’s world view (Schnyder , Buchi, Sensky & Klaghofer, 2000). This has been supported by Karlsson, Berglin and Larsson (2000) who noted a ten per cent increase in patient’s SOC in the year that followed their heart coronary surgery.
Previous research has also noted that increases in SOC is likely following specific psychosocial interventions. Vastamaki, Moser and Paul (2009) implemented a group workshop which practised positive cognitive strategies for a group of unemployed Finnish men. During this time, some of the participants gained employment, the researchers documented the greatest increase in SOC in those who were re-employed. Similarly, Skarsater, Langius, Agren, Haggstrom and Dencker (2005) found that patients who recovered from depression; demonstrated an increase in SOC, where social support was deemed as a significant GRR in their recovery. Szymona (2005) also noted increased levels of SOC following a ten week psychotherapy intervention. Taken together, this literature provides ample evidence that SOC is subject to change following significant life events and effective psychosocial interventions.
SOC is linked with high levels of well-being, self-esteem, self-efficacy and better physical and psychological health (Pallant & Lae, 2002). Past research has implied that greater focus is needed on psychosocial treatment so to strengthen levels of SOC as it is believed to be a good predictor of quality of life (Karlsson, Berglin & Larsson, 2000). Moreover, SOC is associated with a number of physiological and psychological benefits, for example higher levels of SOC are found in people who exercise regularly and feel more socially connected to others (Hassmen, Koivula & Uutela, 2000). It has been shown that people who score high in SOC, in comparison to those who score low are more motivated to adapt to situations, appraise events as challenges as opposed to threats and believe that they have sufficient resources to manage (Eriksson & Lindstrom, 2006). People with high levels of SOC are also more likely to employ suitable coping mechanisms so to alleviate levels of stress (Pallant & Lae, 2002). Additionally, it has been shown that high levels of SOC are negatively associated with depressed mood, stress and anxiety (Karlsson, Berglin & Larsson, 2000). A study conducted by Schnyder et al., (2000) found that patients who had experienced severe accidents and rheumatoid arthritis expressed less anxiety and symptoms of depression when they reported a strong SOC.
The fundamentals underlying this study are supported by Soderberg, Lundman and Norberg (1997) who investigated the effects of SOC in relation to people who suffer with Fibromyalgia; where one experiences chronic muscle and connective tissue pain. Their study found that those who reported a strong SOC experienced higher levels of well-being as they recognised more meaning within their daily life, despite their painful illness. This highlights that people who score high in SOC in comparison to those who score low are more inclined to identify meaning within their life and recognise positive aspects of daily events. This study provides an explanatory basis for the current study as Seligman et al., (2005) encouraged people to identify and report positive events in their daily life.
The goal of the current research is to support and expand upon the findings of Seligman et al., (2005). It is expected that people who are asked to write about three good things in their day and the reasons for these will report greater increases in happiness and SOC, compared to those asked to write about pleasant events from their childhood. The researchers hope to validate the findings of Seligman et al., (2005) by applying this intervention to an Irish population, where it will be determined if this PPI is effective in a cross-cultural setting. Further, this research aims to show that SOC mediates this increase in well-being. A secondary aim is to investigate if frequency of writing during the intervention will lead to a greater increase in well-being. Seligman et al., (2005) did not record the frequency of writing responses, yet asked participants if they continued to write following the intervention. It was found that people indicated that they continued to write for longer period of time had the greatest increase in well-being. It is thought that people who practise PPI’s for longer period of time have “greater opportunity to turn activities in to long lasting habits” (Sin, Della Porta & Lyubomirsky, p.88, 2010). It is also expected that those who write about three good things in their day will instigate more meaning when describing these events in comparison to those who recall pleasant childhood memories. Finally, participants who recall three good things in their day will associate this task more with how they define themselves rather than those who recall pleasant childhood memories.
A total of 67 undergraduate students and staff (37 female, 30 male) from the University of Limerick participated. The age of the sample ranged from 18 to 59 years (M=26.10, SD=10.06). Their socio-economic status was measured indirectly through parent/spouses occupation’s using the Registrar General’s social class classification which is based on occupation as a measure of socio-economic status (Office of Population Censuses and Surveys, 1978). In total, 22.4% reported themselves or their parents as professionals (e.g. doctor, lawyer); 25.4% managerial status; 10.4% skilled non-manual professions, 28.4% skilled craftsman, 7.5% as partly skilled and 4.5% as unskilled labourers. Participants were recruited in the university campus where e-mails were sent to all members of staff and students. An information sheet describing the study and inclusion criteria (over 18 years, ability to speak fluent English) was sent to over 100 people who expressed interest (see appendix A). Following this, the above 67 confirmed participation and the remaining 33 were not included on the basis of English not being their primary language or not having the time to commit to the study’s protocol. All participants gave informed consent, after the study received approval from the U.L EHS Ethics board (FYP 10-009), (see appendix B for Ethics confirmation).
A one-way repeated measures design was employed. The between subjects factor was group (intervention/control groups) and the within subjects factor was time which was assessed over three time periods, baseline (T1), immediately after the intervention (T2) and six weeks later (T3), time was also the independent variable of the study. The dependent variables were changes in happiness as well as SOC which was included as a DV to test the second hypothesis. Furthermore, SOC and frequency of writing responses were tested to understand if they were covariates in the main analyses.
Demographic Data. Within the baseline questionnaire (T1), demographics were collected including questions such as age, gender, ethnicity, martial status, parental/spouse’s occupation and an email address that the participants used most regularly. The research also noted the dates of each time point when the questionnaires were completed.
Authentic Happiness Inventory. The Authentic Happiness Inventory was used to measure one’s current state of happiness (see appendix D). The 24-item scale asked participants to indicate their response on a five point likert scale varying from A, a negative statement (I am usually in a bad mood) to E a positive statement (I am usually in an unbelievably great mood), there was no reversed items in this scale. The scale asked participants to specify the statement that was most relevant to them, in the past week including their feelings at present. In T3, people were asked to indicate their well-being according to feelings over the past six weeks, including their feelings at present. A high score on the scale indicated that one is happy while a low score indicated that one is not happy. The validity of this scale has been supported by Seligman et al., (2005) who compared it to fellow happiness scales such as the Steen Happiness Index and General Happiness Scale. The scales reliability has also been supported in previous studies (Shiffrin & Nelson, 2008; Seligman et al., 2005; Siberman, 2007), within this study, cronbach I± = 0.91.
Orientation to Life Questionnaire. The 13-item Orientation to Life Questionnaire (SOC-13) was used to determine participant’s level of SOC (appendix D). This scale is comprised of thirteen questions measured on a likert scale, ranging between 1 (never happened) to 7 (always happened). The scale responses are calculated by summing up scores, where one can score between 13 and 91, five of these items are reverse coded. A high score in the scale indicates that one has a strong sense of coherence; a low score in this scale indicates that one has a low sense of coherence. The scale consists of three subscales which underlie Antonovsky’s theory; meaningfulness (e.g. how often do you have the feeling that there’s little meaning in the things you do in your daily life?), comprehensibility (e.g. do you have very mixed-up feelings and ideas?) and manageability (e.g. how often do you have feelings that you’re not sure you can keep under control?) This standardised scale has been tested on different populations, where all have reported strong reliability and validity (Frenz, Carey & Jorgensen, 1993; Ristkari et al., 2009.) Eriksson and Lindstrom (2005) found that within 127 studies while using the SOC-13 scale, the cronbach alpha values range from 0.70 to 0.92. Within this study, the cronbach I± is .91 thus supporting previous studies.
The study was entitled “writing exercises and well-being” in an effort to conceal the positive psychology nature of the study. This ensured that a sample who were not directly interested in improving well-being were included. The intervention consisted of two groups; an experimental group and a control group. Following randomisation of participants using a computer software programme (PASW) and completion of baseline questionnaire measures, participants in the experimental group were asked at the end of each working day to think of three good things in their day (having lunch with a friend, good result in an exam) and then to write them down. The participants were asked to reflect on their active role within the event; for example, texting their friend to meet up, studying for exam. A similar protocol was given to the control group, except they were asked to write about pleasant childhood memories. It was necessary to record these recalled events, as the researchers could assess the personal meaning invoked in describing these events. The intervention also asked participants to contemplate the personal relevance of these recalled events. Specifically, participants were asked how important these events were in regard to defining themselves. Furthermore, all participants received their instructions via e-mail with an excel worksheet, where they could write about the events as well as rate how important the events were in regard to defining themselves. These daily tasks were sent back to the researcher to store but also to monitor the frequency of writing over the two-week period. After the intervention, all participants were informed that they could cease writing and return to the psychology laboratory to complete the follow-up questionnaire.
Participants were employed through recruitment e-mails and snowballing techniques, such as lecture shouts in psychology undergraduate classes and student participation schemes provided class credit to first year psychology students for their involvement. Baseline recruitment continued for a week in a psychology laboratory in the Department of Psychology and all follow-ups were held in the same premises. Participants attended the laboratory and were provided with an information sheet and a consent form that they were asked to sign (see appendix A and C). Participants completed the baseline measure in an environment free from distractions. In order to maintain confidentiality, the participants made up their own identification number (two letters and four numbers) that they would remember; this was used for our statistical analyses. When the data collection was completed the email addresses were deleted. A randomised controlled trial was implemented in this study where participants were randomly allocated to each condition using the specialist software PASW. A manipulation check revealed that the groups were comparable in regard to age, gender and level of happiness and sense of coherence (see results section, table 1). The following week, the intervention was emailed to all participants on a daily basis for the duration of two weeks. The participants were asked to complete the exercise and e-mail back the completed excel worksheet which was stored by the researcher to record for frequency of writing responses. This data also provided explanations for changes in happiness levels through examining the written responses, leading to greater control within the research. After the two-week intervention, all participants were informed that they should cease writing and complete the second questionnaire pack; this was similar to the first but did not contain demographics. The final questionnaire was administered six weeks following the intervention to note any changes in happiness and SOC. All participants were given a debriefing form informing them of the purpose of the study (see appendix E).
The data was analysed using Windows PASW. To assess the success of the randomization procedure, demographic and outcome measures (including SOC, well-being) of both the treatment and control groups were compared at baseline using chi-square and a one-way analysis of variance (ANOVA). The data was tested for outliers using the test shapiro wilk’s to ensure the responses were normally distributed. Means and standard deviations of all measures were gathered using descriptive statistics. A mixed ANOVA was conducted to see if happiness and SOC increased following the intervention. A one-way analysis of covariance (ANCOVA) was conducted so to understand if SOC and frequency of writing responses contributed to changes in happiness. Means and standard deviations of how meaningful the participant found the intervention task were gathered and a one-way ANOVA was employed so to see if there were significant differences between groups. As this was a longitudinal study with multiple follow-ups, a high rate of attrition was a limiting factor in our main analyses, at T2, 37 participants returned and at T3, 26 came back. Therefore, the power of the above tests were significantly reduced which increased our risk of type II error.
The data was tested for outliers at each time point where some of the data showed a non-normal distribution. The shapiro wilk test demonstrated that the responses of the treatment group in T3 SOC were not normally distributed; W = 0.84, p = 0.41, (see table I). However, this can be attributed to the low sample size in this condition as there were only ten participants in this group. Manipulation checks revealed that groups were successfully randomised at baseline where no significant differences were noted between groups on the main variables of interest including socio-demographics (see Table II).
Manipulation Check of all Measures at Baseline.
X2 = 0.34, p = 0.84.
F = 0.65, p = 0.89.
Sense of Coherence
F = 0.92, p = 0.60.
Authentic Happiness Inventory
F = 0.96, p = 0.54.
Table III shows the means and standard deviations of the variables-Happiness and SOC at each time point. The mean scores for happiness increase in the experimental condition at T2 and then decrease at T3. Whereas control group (pleasant events from childhood) mean scores decrease at T2 and then increase at T3.
Means and Standard Deviations of Happiness and SOC at each time point.
Three good things
M = 75.88,
SD = 10.51.
M = 77.22,
SD = 13.68.
M = 74.89,
Pleasant Childhood Memories
M = 77.21,
SD = 13.65.
M = 76.64,
SD = 15.38.
M = 83.14,
SD = 21.22.
Three good things
M = 62.11,
SD = 10.52.
M = 61.77,
SD = 10.81
M = 62.88,
SD = 13.95.
Pleasant Childhood Memories
M = 58.93,
SD = 14.39.
M = 61.43,
SD = 14.27.
M = 61.56,
SD = 12.77.
Does Happiness increase following a PPI?
To identify if happiness significantly increased over time, a mixed design ANOVA was conducted, where no significant main effect was found between time (IV) and happiness (DV) in both the treatment and control conditions (F(1, 21) = 0.36, p = 0.69, partial I·2 = 0.04). There were no significant interaction effects found between time-points and happiness (F(1, 21) = 1.18, p = 0.33, partial I·2 = 0.11). Also it was shown that there were no significant differences in happiness found between the treatment and control group (F(1, 21) = 0.24, p = 0.63, partial I·2= 0.01), which can be seen in figure I. Within this ANOVA a low statistical power was noted (within groups = 0.23, between groups = 0.08).
Does SOC increase following a PPI?
To understand if SOC increased over time, a mixed design ANOVA was conducted, where no significant main effect was found between time (IV) and SOC (DV) in both the treatment and control conditions (F(1, 23) = 0.22, p = 0.80, partial I·2 = 0.02). There also was no significant interaction effects found between time and SOC (F(1, 23) = 0.38, p = 0.69, partial I·2 = 0.03). No significant differences was found in SOC between treatment and control group (F(1, 21) = 0.09, p = 0.77, partial I·2 = 0.01) which can be seen in figure II. Within this ANOVA a low statistical power was noted (within groups = 0.10, between groups = 0.06).
Did SOC contribute to changes in Happiness?
A one-way ANCOVA was conducted to determine if SOC accounted for any variance in changes in happiness across three time points. SOC did not account for any significant variance in happiness between T1 and T2 (F(1, 20) = 0.22, p = 0.65, partial I·2 = 0.01) or between T2 and T3 (F(1, 19) = 0.12, p = 0.73, partial I·2 = 0.01).
Did frequency of writing contribute to changes in Happiness?
A one-way ANCOVA was conducted to determine if frequency of writing the interventions tasks mediated changes in happiness across three time points. This analysis was found to be not significant (F(1, 20) = 0.40, p = 0.54, partial I·2 = 0.02).
Meaning derived from the daily PPI task.
During the intervention, all participants were asked to rate how important the written events were in regard to defining themselves. Participants who recalled pleasant childhood events condition scored higher (M = 3.32, SD = 0.85) than participants who recalled three good things a day (M = 3.68, SD = 1.03). However, a one-way ANOVA demonstrated that these differences were not statistically significant (F(1, 22) = 0.87, p = 0.36, partial I·2= 0.04).
Within this study, the findings of Seligman et al., (2005) were not supported as no significant increase in happiness was noted among participants who recalled three good things in their day, in comparison to those who recalled pleasant childhood memories. Additionally, this study fails to reject its null hypothesis as no significant increase in SOC was noted in those who recalled three good things in their day. This research cannot conclude if