1 Introduction

For a long time, psychologists regarded old age as an essentially unhappy time of life, characterized by worsening health, deterioration in cognitive abilities, impoverishment of emotional life and social isolation (e.g., Cumming and Henry 1961). However, over the last few years, a more optimistic view has taken the place of this negative view of old age. As of 1990, an increasing number of psychogerontologists began considering aging as a process of adaptation, requiring of each individual the adoption of strategies for tackling the physical and psychological changes associated with advanced age (Baltes and Baltes 1990; Baltes and Carstensen 1996; Rowe and Kahn 1998). More recently, other authors have gone so far as to describe old age as a wholly positive stage of life, in which an individual can acquire new knowledge, seek pleasant experiences, realize his/her potential and thus achieve high levels of subjective well-being and life satisfaction (Fernández-Ballesteros et al. 2007; Gergen and Gergen 2003; Hill 2005; Ranzijn 2002; Williamson 2002). Note that the concepts subjective well-being and life satisfaction are used interchangeably in current literature and are often likened to the concept of happiness. In fact, all three of these concepts refer to an overall judgment about the quality of one’s own life, which encompasses both affective and cognitive aspects (Diener 2000; Diener et al. 2003; Triandis 2000; Uchida et al. 2004; Veenhoven 2000).

In empirical studies on happiness and subjective well-being in old age, there is a considerable body of research concerned with identifying the objective, personal, and social factors which favor or hinder the achievement of high levels of happiness. These studies examined populations of older adults living in countries that are often very different from each other, although without making comparisons between them. Overall, the results of these studies showed that happiness in older adults is significantly influenced by several factors which crop up repeatedly, such as their physical health status (Girzadas et al. 1993; Ho et al. 1995; Michalos et al. 2000; Smith et al. 2002), the quality of their relationship with family members and, more generally, the presence of sources of social support (Fernández-Ballesteros 2002; Kim et al. 2000; Krause et al. 1992; Lang and Schütze 2002; Philips et al. 2008; Thompson and Heller 1990; Yeung and Fung 2007).

In order to gain information on the role played by socio-cultural factors in the aging process, other studies investigated the levels of happiness reported by the older adults and the factors which influence them, by way of cross-national research designs. Indeed, there are still few studies conducted in this field, although this trend is gradually changing for the better (Antonucci et al. 2002; Walker and Lowenstein 2009). In a recent study, Fagerström et al. (2007) investigated life satisfaction among people aged 60–89 years from six European countries (The Netherlands, Luxemburg, Italy, Austria, United Kingdom, and Sweden). The results showed that most of the respondents reported being satisfied with their lives; however, the Italians scored a lower median level of life satisfaction than participants from other countries. The authors also found that factors contributing to life satisfaction were similar in all surveyed countries; these factors included satisfaction with social contacts, financial resources, subjective perception of health problems, and self-esteem. In another study, Katz (2009; see also, Lowenstein et al. 2007) examined the effects of patterns of family relations (solidarity, conflict, and ambivalence) on the subjective well-being of older parents from five countries (Norway, England, Germany, Spain, and Israel). Three a priori indicators of subjective well-being were used in this study: Life satisfaction, and positive and negative affects. Findings indicated that Israelis reported lower levels of subjective well-being in comparison with other participants, except for positive affect. Furthermore, the extent of family solidarity turned out to be a significant predictor of all three of the components of subjective well-being, in all the countries. This confirms the role played by family support in promoting happiness in old age.

A very limited number of cross-national studies have also compared happiness of older adults living in countries with an individualist versus a collectivist cultural orientation. The rationale behind this comparison is that these two orientations involve opposing ways of understanding the role of the individual within society (cf. Kim et al. 1994; Triandis 2001). This could have a significant effect on the strategies by which older adults cope with old age and on the levels of happiness they can achieve. However, it should be noted that the results obtained by this research line are discordant, indicating similarities between individualist and collectivist cultures in some cases and marked differences in others. Silverman et al. (2000), for example, investigated the relationship between social networks and life satisfaction in older adults from the United States and China, respectively. Overall, results showed that the American participants reported greater life satisfaction compared to the Chinese participants. The cultural context also differentially influenced how the structure of social network related to the life satisfaction scores. In another study, Venkatraman (1995) tested a model of factors influencing subjective well-being in two samples of married older adults from the United States and India. Results showed that the mean differences for subjective well-being were not statistically significant across the two samples. Moreover, Americans and Indians did not differ in terms of the influence of emotional support from role relationships (e.g., adult children, spouse, friend) on their subjective well-being.

In summary, the above studies highlighted that happiness in older adults is influenced by a set of diverse factors. Of these, an important role is played by the individual’s health, the quality of family relationships and, albeit to a lesser degree, the cultural context. It is important to point out that although these studies have enabled researchers to gain valuable information on the objective, personal and social conditions which influence quality of life in old age, they share at least two important limits. The first limit concerns the ways the researchers measured happiness. All the studies used questionnaires in which the older adults had to express their agreement with a list of statements or to quantify their level of life satisfaction by way of a numerical scale (e.g., 1 = dissatisfied, 10 = satisfied). It should be noted that the researchers who used this approach took it for granted that all the participants in their studies shared a common idea of happiness. However, this assumption could be unjustified in that different people could assign different meanings to the idea of happiness and thus provide assessments of their happiness levels which prove to be incomparable, being based on different subjective representations of this concept. This problem becomes particularly significant when comparisons are made between the judgments of individuals coming from different cultural and linguistic backgrounds, in which the same words happiness and well-being could have profoundly different meanings or may not even be translatable from one language to another (see Lu and Gilmour 2004; Wierzbicka 1999).

A second limit to the studies conducted to date concerns the researchers’ choice of factors which, in their opinion, influence the participants’ levels of happiness. Indeed, the factors considered in the different studies were selected on the basis of theoretical models of aging, but without taking into consideration the naïve representations of happiness formulated by the participants and therefore without verifying in advance whether they considered these variables to be important components of their concept of happiness.

From the above, it appears that in order to obviate all these problems, it is necessary to first collect data on the older adults’ naïve representations of happiness formulated individually, without providing suggestions to the research participants, and only then to measure the degree to which they consider they have realized these ideal representations in their life. This approach is particularly necessary when conducting cross-cultural studies and, if applied correctly, would enable researchers to gather information on the cultural differences in the way of conceiving happiness, while being able to make more reliable comparisons between the opinions expressed by the individuals on their level of subjective well-being. Indeed, Galati et al. (2006a) have already used this approach in a cross-cultural study aimed at exploring the naïve representations of happiness put forward by the Italians and Cubans; however, the data presented in that study refer to a sample consisting mainly of young people and adults (mean age = 32.5 years) so do not provide specific information relating to older adults.

1.1 The Present Study: Research Contexts, Aims and Hypotheses

This study extends the previous work by Galati et al. (2006a) by investigating the subjective representation of happiness and its level of attainment in two groups of healthy older adults from Italy and Cuba. Italy and Cuba were chosen in the present study as well, since they present a set of very different ethnographic, socio-economic, and cultural characteristics, some of which are considered by contemporary literature to be factors that are capable of differentially influencing happiness (Lu and Gilmour 2004; Triandis 2000; Uchida et al. 2004). We shall describe below the main characteristics of the cultural contexts under investigation.

Italy is a European country inhabited by a Caucasian population; Cuba, on the other hand, is a Caribbean country whose population is a mix of ethnic groups from the Iberian Peninsula which settled on the island during the Spanish colonization of America (Cayuela Fernández 1993), and ethnic groups from Africa, shipped to the country during the time of slavery. From a cultural standpoint, the dominant religion in Italy is Roman Catholicism, whereas Cuba epitomizes a synthesis between the Christian and the African animist traditions. Considering the economy and the socio-political organization, Italy is a society with a capitalist-type market economy; Cuba, on the other hand, is a developing society with an economy linked to the socialist model (Le Riverend 1985). Different levels of socio-economic development between Italy and Cuba emerge when the Human Development Index (HDI) of these countries is taken into consideration. This index takes into account longevity, knowledge (rate of literacy and school population) and standard of living (Gross Domestic Product per capita, GDP) and is generally proposed as the best measure of the development of a nation (Cordelier and Didiot 1997). According to recent statistics, Italy’s HDI is 0.93 and is similar to that of other highly developed countries, such as New Zealand and Germany. Cuba’s HDI is 0.82, which is closer to that of other developing countries, such as Mexico and Bulgaria (NationMaster 2008). Interestingly, the differences between Cuba’s and Italy’s HDI are due mainly to the differences in their income levels: GDP per person in Italy is 30,073 dollars, which is more than ten times greater than the GDP per person in Cuba (2,863 dollars). However, Italy and Cuba do not differ significantly in terms of longevity and literacy: The percentage of literate persons out of the whole population is 98.4% in Italy and 99.8% in Cuba, while life expectancy is 79.9 years in Italy and 77.1 years in Cuba.

Italy and Cuba also differ significantly with regard to the role attributed to the individual and to the group within their respective cultures. Italy has an essentially individualistic cultural orientation, whereas Cuba is characterized by a collectivist-type structure. Basabe et al. (2002) reported the ratings of individualism for 54 nations, which included Italy and Cuba: Italy had a score of 76 out of 100, whereas Cuba’s score was only 12. The less individualistic character of Cuban culture is also confirmed by a recent cross-cultural study on the everyday emotional life of young Cuban and Italian adults (Galati et al. 2004, 2005). The results of this study revealed that the emotional antecedents relating to goal attainment and success were cited rarely by the Cuban participants and were not associated with positive emotions, as opposed to what was found among the Italian participants. Moreover, Cubans reported sharing their emotional experiences most often with friends and the family, showing a leaning towards the little group and collectivity (cf. Galati et al. 2004).

Bearing in mind the socio-cultural characteristics of the two countries under investigation, we designed a questionnaire study aimed at tapping the cross-cultural similarities and differences between the subjective representations of happiness reported by participants from the two countries and their levels of attainment. Specifically, the research set out to achieve the following objectives:

  1. (1)

    to identify the specific components which characterize the subjective representation of happiness in the two samples under investigation, and compare the prevalence of the components and their relative importance;

  2. (2)

    to measure the level of attainment of the different components of happiness reported in the two countries and, starting from these judgments, to find an overall happiness indicator for each cultural group, based on the different representations of happiness put forward by each participant in the study;

  3. (3)

    to compare the overall level of happiness attained in the two cultural contexts under investigation;

  4. (4)

    to identify the factors which influence the overall level of happiness attained by the study participants, considering participation in social and cultural activities, the participants’ perceptions about their transition into old age, religious beliefs, and socio-demographic characteristics, as possible explanatory variables besides the cultural context to which they belong.

To attain these objectives, we used individual self-reports, including open-ended questions to reveal cultural nuances which are liable to be masked by scales including pre-determined components. The main hypothesis was that the cultural context might influence the representation of happiness on both a qualitative level (regarding the subjective representation of components and their importance) and a quantitative level (regarding the level of attained happiness components).

2 Method

2.1 Participants

Participants in the Italian and Cuban groups were recruited from the cities of Turin (Northern Italy) and Havana respectively. In order to make comparisons between the two groups, participants from the two countries were selected according to a similar criterion. More specifically, the total sample of each country was stratified by age, gender, educational level, and three different ways of participating in social and cultural activities. The first group included people who habitually go to social centers and meeting places for older adults where it is possible to take part in games, as well as people who carry out voluntary work or sporting activities in various kinds of public and private organizations (socialized participants). The second group included participants who are involved in cultural activities, such as University Programs for Older Adults or courses, events and initiatives organized by associations and cultural centers in their area of residence (participants involved in cultural activities). Lastly, the third group contained older adults who do not take part in any social or cultural activity and who, therefore, lead a rather solitary life (not socialized participants).

The total sample consisted of 395 subjects aged between 60 and 97 years: 209 lived in Italy (36.4% males, 63.6% females, age 60–91 years) and 186 lived in Cuba (27.4% males, 72.6% females, age 60–97 years). Table 1 presents the socio-demographic characteristics of the study participants from the two cultural groups. Bivariate analyses (chi-square and t test) confirmed that the two groups were comparable with respect to variables used to pair them (i.e., age, gender, educational level, and participation in social and cultural activities). However, there were substantial differences in marital status across the two samples: There were more married people among the Italian respondents, whereas the Cuban group included a greater proportion of separated/divorced respondents; lastly, there were no widowers in the Cuban group.

Table 1 Socio-demographic characteristics of the study participants
Full size table

2.2 Procedure

Participants in the study were recruited through cooperation with organizations and institutions in the cities of Havana and Turin. More specifically, participants belonging to the group of older adults who were socially active and involved in cultural activities were contacted with the assistance of social centers, sports and cultural associations, and voluntary organizations within the different areas of the cities considered. To recruit older adults who were not socially active, we followed a number of leads provided by members of the first two groups (socialized and involved in cultural activities), who indicated people they knew who were not involved in any social or cultural activity of the kind described above, and who therefore did not belong to any cultural, sports or voluntary organization. The social condition and lifestyles of all the participants were further investigated through an interview carried out by the authors of this study during the selection of the participants and this enabled us to verify whether the subjects had been allocated to the three research groups correctly. Importantly, this interview also enabled us to identify certain subjects who were experiencing serious socio-economic hardship or whose health conditions were invalidating, thus allowing us to exclude marginal cases from both samples.

All participants contacted via the above-mentioned channels were told that the University of Turin and the University of Havana were conducting a study on things that make people happy. Those who agreed to participate were then asked to fill out a questionnaire. Participants were not paid and took part in the investigation voluntarily. Respondents were also informed that the questionnaire was anonymous and that their answers would be used for research purposes only.

2.3 Instruments

We administered a modified version of the questionnaire used by Galati et al. (2006a) in their previous investigations on happiness conducted on Italian and Cuban participants belonging to various age groups (see also Galati et al. 2006b).

The questionnaire employed in the present study had three sections. The first section was more or less identical to the first section of the original questionnaire by Galati et al. (2006a) and it aimed to investigate the subjective representation of the components of happiness and their level of attainment. The second section was created by the authors of this work specifically for this study and aims to investigate the aspects of the individual’s personal life that are linked to the transition into old age and adjustment to this stage of life. Lastly, the third section identifies the participants’ main socio-demographic characteristics. The items included in the different sections of the questionnaire are described in detail below (the first and second section of the questionnaire are reported in Appendix).

2.3.1 First Section

Participants were asked to think of happiness and to write down at least five things that made them feel happy (“happiness components”). They were asked to record the happiness components in order of importance. To measure the level of attainment of each of these components, respondents were then asked to evaluate on a 5-point rating scale to what extent they had attained them in their lives (0 = not at all, 1 = a little, 2 = moderately, 3 = quite a lot, 4 = totally).Footnote 1 This part of the questionnaire was designed with reference to the theories of happiness based on discrepancy measures (Michalos 1985; Parducci 1995). Cronbach’s alpha coefficients for the attainment scores were .53 in the Italian sample and .56 in the Cuban sample. These coefficients are quite low, suggesting that participants viewed the degree of attainment of their happiness components as varying considerably. Moreover, the attainment scores calculated for the various components are assumed to be formative indicators (see Bollen 1989) of the overall happiness index we used in the present study; this implies that their covariations must not be necessarily high.

2.3.2 Second Section

Four closed-ended questions asked participants to judge whether they currently considered themselves useful (yes/no), with responsibilities (yes/no), active (yes/no) and as having aspirations (yes/no).

2.3.3 Third Section

We assessed age, marital status, education (scored both as levels of education and as years of study), and nationality. Furthermore, participants were asked whether they had religious beliefs (yes/no).

The original version of the questionnaire was composed in Spanish and then translated into Italian. A back-translation was performed by two bilingual speakers not involved in the study. A few differences appeared and the two translators agreed on how to resolve them. The final version of the questionnaire was devised after a pilot investigation, conducted in both Italy and Cuba, which enabled us to improve the formulation of the questions. On average, filling out the questionnaire took 20 min.

3 Results

3.1 The Nature of Happiness Components

Overall, the participants in the research cited between 1 and 10 happiness components, the mean being 4.66 (SD = 1.31). This indicates that not all the participants were able to comply with the item in the questionnaire which asked them to write down “at least five things” that made them feel happy. The comparison between the two cultural groups revealed that the mean of components cited was significantly higher in the Italian sample (M = 4.97, SD = 1.19) compared with the Cuban sample (M = 4.30, SD = 1.36), t(371) = 5.18, p < .001.

In order to identify the nature and the meaning of the happiness components, the answers provided by the participants were grouped into categories. The categorization was performed by the authors of the present work on the basis of the criterion of semantic similarity. For example, the answers “harmony amongst the members of my family, “looking after grandchildren”, “spending a lot of time with my children” were included in the category family; “more social justice”, “peace in the world”, “equality” were classed in the category values; “traveling”, “seeing a good film”, “entertainment” were included in the category pleasant events. The categorization work was performed by collapsing all the answers given by the participants from the two groups taken together. This procedure allowed us to identify 25 categories of happiness components. Five categories were culture-specific, but they were cited by a minority of participants from the two cultural groups. Security and dealing with adversities were cited solely by the Cuban participants (respectively 7.6 and 5.4% of this group reported them); in turn, the Italian participants cited hobbies (12.4%), sex (1.4%), and good luck (1.0%). The remaining 20 categories were shared by the two cultural contexts examined.

Given the low frequency of the culture-specific components, we decided it was simpler to compare the two groups of participants by considering only the 20 categories shared by both cultural groups. With regards to the semantic content of these categories, seven of them refer to individual interests, ranging from the satisfaction of primary needs (health, money, home) to higher motivation, such as self-realization and the desire to learn (work, success, self-actualization, culture and knowledge); five components refer to the emotional dimension (pleasant events, absence of unpleasant events, positive emotions, empathic emotions, serenity/well-being); a further five components concern the person’s relational interests. Of these components, three refer to relationships with specific persons (family, friendship, partner), while the other two have a more general content (love, good affective relationships). More in detail, love refers to an abstract concept that is very close to the Christian concept of brotherly love; good affective relationships, on the other hand, refers to relationships with various kinds of persons belonging to the individual’s social network, such as neighbors and acquaintances. Finally, the three other components we identified in this study refer to moral and value-related aspects linked to realizing a project for a happy life (values, faith, altruism).

In order to verify the validity of the happiness components’ categories, we asked six final-year psychology students (three Italians and three Cubans)—who were unaware of the aims of the study—to codify the responses provided by the two cultural groups, using the 25 categories identified in the first stage of the analysis. The correspondence between the codification performed by the authors of the study and the one by each of the naïve judges was then calculated. These comparisons were made two by two, using Cohen’s k. All k values were greater than .7 (p < .05 in all cases), thus confirming the validity of the categories.

3.2 Frequency and Rank of Happiness Components

In order to describe the representation of happiness in the two groups of participants, we computed the frequency and the rank order of citation of the 20 happiness components shared by the two countries. Table 2 presents the entire list of categories of happiness components in order of frequency of citation (calculated as the percentage of participants who mentioned them) in the Italian and the Cuban sample. As can be seen, the categories cited by more than 50% of participants were health, family, and money in Italy, whereas only health reached this percentage in Cuba. Other happiness components frequently reported by the Italians (cited by more than 30% of participants) were, in order, friendship, good affective relationships, and serenity/well-being, whereas the Cubans cited love, faith, good affective relationships, and family. Among the core elements cited by more than 30% of the participants, only three components were shared by the two cultural groups, namely health, family, and good affective relationships. The Italian-specific core elements were money, friendship, and serenity/well-being, whereas the Cuban-specific core elements were love and faith.

Table 2 Frequency of citation and mean rank of happiness components in the two cultural groups
Full size table

Table 2 also presents the mean ranks of happiness components reported in the two samples, which were obtained by calculating the average of the rankings attributed to them by participants. Focusing on the happiness components cited by at least 30% of participants, those considered most important by the Italians were health (M = 1.83, SD = 1.36) and family (M = 2.56, SD = 1.32), whereas the most important components cited by the Cubans were health (M = 1.84, SD = 1.13.) and faith (M = 2.35, SD = 1.30).

In order to verify the existence of a general correspondence between the order of frequency and the order of importance of all the components, Spearman rank-order correlation coefficients were calculated within the two groups of participants. Analyses carried out within the Italian and the Cuban groups showed a positive relationship between the two orderings (rho = .52, p < .05 in the Italian sample, rho = .44, p < .05 in the Cuban sample, N = 20). Taken together, these results show that the components most frequently cited by the two samples corresponded by and large to those indicated with the highest priority, whereas the less frequently cited components corresponded by and large with those ranked low in order of importance. Correlation analyses were also performed on the frequency of citation and mean rank between the Italian and the Cuban sample. The results obtained from both these analyses were not significant (p > .05), thus showing different profiles for the Italian and the Cuban participants.

In order to understand the structure of these different profiles better, we analyzed the relationship between the frequency with which happiness components were cited, their importance, and cultural background (Italian vs Cuban), computing a correspondence analysis of a 20 × 10 cross-tabulation table, in which the 20 rows are the components of happiness, and the 10 columns are the orders given (from 1 to 5) by the Italians and the Cubans. The scores in the cells of this matrix refer to the frequency of citation of a specific component in a specific order in one culture. We choose as normalization procedure the symmetrical method, which distributes the inertia equally over the rows and columns. Figure 1 shows a graph representing the map obtained by crossing the first and the second dimension extracted, which overall reproduce 73.1% of the total inertia. As can be seen, health occupies a distinct, well identified position in the top left hand part of the graph because it was cited very frequently in first place by both the Italians and the Cubans. However, the importance of this component appears to be more significant for the Italian group, given that it is much closer to the first rank position in Italy. Some significant cultural specificities emerge considering the distribution on the graph of the other components in relation to the ranks of citation in Italy and in Cuba. Indeed, two groups appear, specific for the two cultures. In particular, a cluster of components ranked more frequently second to fifth by the Italians appears in the upper right quadrant. The main components of this cluster are, in order of importance, family closer to second place, money closer to third place and following on from these, work, friendship, culture/knowledge, and values. On the other hand, between the lower left and right hand quadrants, there is another cluster which groups together the components cited most frequently by the Cubans. These include love, good affective relationships, and self-actualization, which are closer to the second place. After these, closer to third place, there are positive emotions, absence of unpleasant events and then success and pleasant events, closer to fourth and fifth ranks. The components empathic emotions and faith are the furthest away from all the components cited most frequently by the Italians and thus indicate a particularly accentuated Cuban specificity. Finally, the components placed closer to the graph origin (i.e., serenity/well-being and partner) do not contribute significantly to the characterization of the profiles of the two samples.

Fig. 1
figure 1

Correspondence analysis map of frequency and order of importance of happiness components in the two cultural groups. Note: The symbols n It and n Cu indicate the rank of happiness components given by Italians and Cubans, respectively (from 1 to 5)

Full size image

3.3 The Attainment of Happiness Components

Table 3 presents the mean attainment scores for the most reported happiness components in the two cultural contexts (cited by at least 30% of one of the two samples). The components attained most by the Italians were family (M = 3.31, SD = .82) and good affective relationships (M = 3.19, SD = .79), whereas the least attained component was money (M = 2.43, SD = .97). In the Cuban sample, the components attained most were faith (M = 3.79, SD = .17) and good affective relationships (M = 3.34, SD = .65), whereas the least attained component was money (M = 1.63, SD = .66). Some interesting differences emerged when the level of attainment of the single components was compared in the two countries (Table 3). The results of the Student t test for independent samples showed that the Italians reported higher attainment scores for the family, health, and money components. In addition to these results, a supplementary analysis, performed by considering the components with a low citation frequency as well—and which for this reason was not included in Table 3—showed that the Cubans scored significantly higher in altruism t(47) = 5.15, p < .01), self-actualization t(78) = 2.30, p < .05), and values t(56) = 4.14, p < .01. However, as these components were cited by fewer than 30% of the samples, these data should be verified in future studies conducted on larger samples which would therefore be able to provide more reliable assessments.

Table 3 Mean attainment scores and standard deviations of the most reported happiness components in the Italian and Cuban samples and results of the t test performed to assess group differences
Full size table

3.4 Measuring Overall Happiness

A measure of overall happiness was computed by calculating the mean of the attainment scores attributed by participants to each component. The descriptive analyses showed that the mean scores of overall happiness in the Italian sample and in the Cuban sample were 2.81 (SD = .62) and 2.80 (SD = .66) respectively. Differences in overall happiness were analyzed by means of an analysis of variance, including as between subject factors culture (two levels: Italy and Cuba), the participants’ involvement in social and cultural activities (three levels: Socialized, involved in cultural activities, not socialized), and gender. Results revealed that the only significant factor affecting overall happiness was the participants’ involvement in social and cultural activities, F(2, 394) = 8.40, p < .001. More specifically, the Bonferroni post hoc test (p < .05) showed that participants involved in cultural activities (M = 2.95, SD = .62) were significantly happier than non-socialized participants (M = 2.61, SD = .84), whereas only a marginally significant difference (p = .08) was found between participants involved in cultural activities and socialized participants (M = 2.79, SD = .48). Partial eta squared indicated that 4% of the variance in overall happiness was explained by the participants’ involvement in social and cultural activities.

3.5 Predictors of Overall Happiness

In order to identify the predictors of overall happiness in Italy and Cuba, a linear regression model (stepwise method) was computed, including the independent variables of country (1 = Cuba, 0 = Italy), age (years), sex (1 = male, 0 = female), education (years of study), marital status (1 = married, 0 = unmarried), having religious beliefs (1 = yes, 0 = no), involvement in social and cultural activities (1 = involved, 0 = not involved, with “not socialized” as the reference group), as well as the variables included in the second section of the questionnaire: Responsibility, self-perceived usefulness, aspirations, activeness (all codified as 1 = yes, 0 = no). Since the distribution of marital status significantly varied across the Italian and Cuban samples, we controlled for the possible confounding effect of this variable by including among the predictors of the regression model also the interaction between country and marital status. Results of the regression analysis are presented in Table 4. As can be seen, the best predictor of overall happiness was education (β = .24). Moreover, overall happiness was positively associated with self-perceived usefulness (β = .15) and being married (β = .13). About 11% of the variance in overall happiness was explained by the three significant predictors mentioned above.

Table 4 Significant predictors of overall happiness
Full size table

4 Discussion and Conclusions

The aim of this study was to investigate the subjective representation of the components of happiness and to what extent they are achieved in two groups of older adults from two countries with very different socio-economic characteristics and cultural orientations: Italy and Cuba.

Overall, the results showed important cross-cultural similarities and differences. The analysis of the contents of all the participants’ answers to the open-ended question included in the first part of the questionnaire highlighted that the subjective representation of happiness was based on 25 components. Of these, 20 were cited by both groups, while five were culture-specific. In particular, hobbies, sex, and good luck were cited only by the Italian participants, while security and dealing with adversities were cited by only the Cuban participants. Nevertheless, since all the culture-specific components were cited by a small minority of the groups studied, for clarity and simplicity we decided not to include them in the statistical analysis of data.

With regards to the 20 components that were common to both samples, on which the cross-cultural comparison between the two countries was based, these covered the semantic space of the concept of happiness and referred to diverse spheres of life, experience and interests, including health, money, family, interpersonal relationships and affective experiences and also concerned the areas of values and ethical behavior. It is interesting to note that the list of happiness components identified in this study corresponds largely to the list obtained in previous studies (Galati et al. 2006a, b) conducted on samples of Italian and Cuban participants, consisting of people belonging to younger age groups (young people and adults) than the age group investigated here. Moreover, many of the components identified are seen in another cross-cultural study conducted by Chiasson et al. (1996), who analysed the naïve representation of the factors that make people happy in different groups of young people resident in countries with an individualist cultural orientation (U.S. and Canada) and countries with a collectivist orientation (El Salvador). Furthermore, the components most frequently cited by the participants of the present study were found in several interview surveys conducted in the United Kingdom (Farquhar 1995; Gabriel and Bowling 2004) and Sweden (Wilhemson et al. 2005), which aimed at analyzing older people’s definitions of the more general construct of quality of life.

The frequency of citation and the importance assigned to the different happiness components within the two samples was found to be positively correlated and this orders correlation seems to indicate a hierarchy of typicality. In fact, this kind of correlation has often been found in the studies on the organization of prototypical categories (see, for example, Rosch 1975, 1978). Therefore, it seems reasonable to assume that the study participants’ naïve concept of happiness has the characteristics of a prototypical category. Interestingly, correlation analyses performed on both the frequency of citation and mean rank of happiness components between the Italian and Cuban samples were not significant. These results suggest that, although the representations of happiness in the two cultural groups share some core elements, there are fundamental culture-related differences in the importance attributed to them, which determine clearly different profiles for Italian and Cuban older adults.

A number of interesting indications regarding the cross-cultural similarities and differences in the subjective representation of happiness emerges considering the results of correspondence analysis. This analysis pinpointed that aside from the cultural context, health represents the most significant happiness component (in terms of both citation frequency and importance) for study participants from both samples. However, this result should not be a surprise as good health guarantees that the body functions well, which for an individual, especially an older adult, is a fundamental condition for being able to organize one’s present and future existence and to achieve satisfying levels of subjective well-being. Some significant cross-cultural differences emerge, considering the distribution on the graph of the other components deemed important by the two cultural groups. Specifically, it appears that the Italians attribute a certain importance to components such as family, money, work, friendship, values, and culture/knowledge, whereas the Cubans seem to focus on different components such as love, good affective relationships, self-actualization, positive emotions, absence of unpleasant events, empathic emotions, and faith. To summarize, this seems to indicate that the most typical attributes of happiness in Italians refer more to the sphere of the individual’s interests: the security offered by the family, work as a tool for acquiring money, the acquisition of knowledge as an effective tool for acting in the environment and, lastly, a system of rules guiding the individual’s behavior. On the other hand, the prototypical attributes of happiness in Cubans seem to be less centered on the individual and refer more to the emotional, relationship, and communication aspects of life, thus placing the social dimension of existence at the forefront.

These cross-cultural differences in the subjective representation of happiness components are confirmed by the results obtained by analytically comparing the attainment scores assigned to them by the participants. A brief look at the results shows that the Italians believed that they had achieved health, family, and money to a greater extent than their Cuban counterparts. On the other hand, the Cubans scored significantly higher in components such as altruism and self-actualization. On the basis of this evidence, the individualist orientation of the Italian cultural context and the collectivist orientation of the Cuban cultural context are confirmed.

Further indications of factors which can affect subjective well-being may be gained by considering the overall happiness of the study participants. This indicator was built on the basis of the attainment scores assigned to individual happiness components and may be considered as an indicator which can provide reliable information on the participants’ level of life satisfaction and subjective well-being. It is important to note that cross-cultural differences disappear at this level of analysis. Indeed, the level of overall happiness was not significantly different in the two cultural groups examined. This result suggests that the presence of high levels of material well-being and income, such as those found in Italian society, are not a sufficient condition for reaching higher levels of subjective well-being than those observed in developing countries. In other words, the older adults living in Cuba, despite having no access to many economic resources, report happiness levels that are comparable to those of their Italian counterparts, and this confirms the thesis put forward by several authors that there is no direct relationship between the levels of subjective well-being and the levels of economic well-being observed in a given country (Csikszentmihalyi 1999; Galati et al. 2006a; Myers 2000).

Although culture and the country of origin do not have a significant impact on the levels of overall happiness of the participants, the analysis of the data showed that overall happiness varied significantly according to participation in cultural activities. In particular, it has been found that the fact of attending University Programs for Older Adults or taking part in initiatives organized by cultural associations was related to rather high levels of overall happiness. On the other hand, older adults living in relative isolation had lower scores for overall happiness than participants who took part in either cultural or social activities. These results show that possessing suitable cultural tools and the presence of a consolidated social network are a fundamental resource for guaranteeing a good quality of life in old age. By contrast, it is well documented that living alone or in isolation are risk factors that increase the likelihood of the older person experiencing depressive symptoms (Chi and Chou 2001; Dean et al. 1992; Ramos and Wilmoth 2003).

The importance of individual cultural resources as a tool for improving the experience of old age has also been confirmed by the results of the regression analysis, which showed a positive relationship between the number of years of study and the level of overall happiness. This analysis also indicated that high levels of overall happiness were associated with participants’ perception of themselves as useful persons. This information confirms findings from previous studies (Fernández-Ballesteros 2002; Kim et al. 2000; Krause et al. 1992) showing that providing support to other people and, more generally, self-perceived social efficacy play an important role in enhancing subjective well-being in old age. In line with previous research conducted on various age groups resident in countries with different cultural orientations (Gove et al. 1983; Haring-Hidore et al. 1985; Inglehart 1990), the regression analysis also highlighted how overall happiness is greater in older adults who are married. To interpret this result, we can hypothesize that the existence of an affective relationship providing constant social and material support is a facilitating factor for facing the challenges and tasks that characterize old age. Lastly, it is worth noting that the regression analysis had lower predictive power. This suggests that predictors other the ones considered in the present study could better explain variance in overall happiness.

Naturally, the data presented here cannot be considered conclusive as the samples investigated were limited in size and the employed instruments could be improved. Future research could examine larger groups of older adults, by using self-report instruments including open-ended questions and scales on happiness components (like the ones used in the present study), as well as well-validated measures of subjective well-being (e.g., Diener and Diener 1995; Diener and Suh 2000). This would allow researchers to compare the results of investigations about the naïve representation of happiness with other studies in the field of happiness and subjective well-being. Furthermore, it would be important to collect data on a wider spectrum of personal and social factors that could influence happiness of older adults living in different cultures; among these, of particular interest are the size and structure of social network and the subjective perception of physical health status. Finally, for a better understanding of the impact of age on the subjective representation of happiness and on its levels of attainment, it would also be useful for the researchers to compare the representations and opinions of the older adults with those of the younger groups. This might enable us to identify the key transitions and critical moments in the development of the subjective representation of happiness over the life cycle. Furthermore, these data could be used to prepare preventive intervention strategies on an individual and a social level.