According to Veenhoven (1984), happiness is “the degree to which an individual judges the overall quality of his own life as-a-whole favorably” (p. 22). In other words, happiness expresses how well an individual likes the life he or she leads. This evaluation is based on the sensory system, cognition and affect. Happiness in general includes equally one’s quality of life in the past, presence and future. On the other hand, happiness is not necessarily stable: people may change their attitudes towards life over time (Grant, Button & Noseworthy, 1994). Specific aspects of happiness like satisfaction with different life domains can contribute to happiness as a whole. These aspects of life such as satisfaction with work or family do not say obligatory something about the general livability of a society. Citizens can be satisfied with their work, but still be unhappy because their society offers little more (Veenhoven, 1993). Also to consider is the finding that aspects of life are not equally important in all societies at all times. Work for instance is less central in most third world countries than in the western society (Cherns, 1984). For this reason, when comparing different countries, it is more informative to focus on happiness or satisfaction with life as a whole (Veenhoven. 1993).

Item generation and selection

The items of the Happiness and Satisfaction Scale were used in the module ‘Family and Changing Gender Roles’ of the ISSP in 2002 and 2012. It was included because the items were supposed to be useful for understanding and evaluating outcomes among different types of family conditions, e.g. for work-family conflicts (Scholz, Jutz, Edlund, Oun, & Braun, 2014). Item 1 measures happiness in general, item 2 measures satisfaction at work and item 3 satisfaction with family life. The latter two items were specifically developed for the family and gender related topic by the German ISSP drafting group. Each of the three items include opposite dimensions of happiness and satisfaction – completely happy vs. completely unhappy (item 1) and completely satisfied vs. completely dissatisfied (item 2, item 3). The mean score of all three items can be used to examine happiness and satisfaction in different life domains, whereas the single items can be used as indicators for happiness and satisfaction in specific life domains, for example to examine the relationship between the presence of work-family conflict and the degree of happiness and satisfaction (Grönlund and Öun, 2010).

Samples

The findings reported below are based on the samples of the International Social Survey Programme (ISSP) 2012. Participants were randomly selected from all persons aged 15 and over resident within private households. In some countries, participants were selected regardless of their nationality. The following countries completed the Happiness and Satisfaction Scale: Argentina, Australia, Austria, Bulgaria, Canada, Chile, China, Croatia, Czech Republic, Denmark, Finland, France, Germany, Great Britain, Iceland, India, Ireland, Israel, Japan, South Korea, Latvia, Lithuania, Mexico, Norway, Philippines, Poland, Russia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Taiwan, Turkey, United States, Venezuela.

The sampling procedures differed between countries. The sampling procedure for each country is described in the ISSP 2012 variable report. The final sample consists of N = 56,254 people from 37 countries, including N = 26,039 male and N = 30,151 female respondents. The average age is 47.59 years (SD = .50).

For the present analyses, the answer categories „Can’t choose”, “No answer”, “Doesn’t apply, no job” for item 2 and “Not applicable, no partner” for item 3 were classified as missing values. Furthermore, many countries provided design weights to correct for different sampling probabilities in the countries. These weights were applied in the following analyses to represent all country subsamples equally. See Table 2 for sample size, age, gender and educational level for each country.

Table 2

Sample size (N), gender (%), age (M, SD) and educational level (%) for each country and across all countries (total)

Note. Observations were weighted based on design weights. The gender and educational level data do not sum up to 100% because the categories „no answer” and “don’t know” are not reported. ^*1 = no formal education, primary/elementary school or lower secondary, 2 = upper secondary or post-secondary (allows entry university, other programs toward labor market or technical formation), 3 = tertiary level (Bachelor, Master, doctoral degree).

Item analyses

We used a multi-group structural equation model to investigate the dimensionality of the scale. Parameters were estimated using the robust maximum likelihood estimator in Mplus. The configural invariant, tau-congeneric model with one latent variable, three manifest variables and different factor loadings is exactly identified and the model fit cannot be tested (df = 0). Therefore, we examined a metric invariant, tau-congeneric measurement model where the factor loadings were constrained to be invariant across countries, whereas the variance of the latent variable and the items’ intercepts were allowed to differ. This model fits the data well (RMSEA = .049, CFI = .980, χ²(70) = 317.296, p = .000). Additionally, we tested a metric invariant, tau-equivalent model with equal factor loadings for all items, which did not lead to an enhancement of the model fit (RMSEA = .162, CFI = .778, χ²(72) = 2,883.573, p = .000). Therefore, we used the metric invariant, tau-congeneric model to estimate reliability. The model structure with standardized regression weights for Germany is presented in Figure 1.

Figure 1. HAS = Happiness and Satisfaction, structure and standardized parameters of the model, RMSEA = .049, CFI = .980, χ²(70) = 317.296, p = .000, N = 1,763.

Item parameter

Means and standard deviations for Germany and for the total sample are shown in Table 3. See Table 7 for reference values for each country.

Table 3

Mean and standard deviation of the manifest items for Germany and across all countries (total)

Note. Observations were weighted based on design weights, inverted scale from 1 = “completely unhappy” to 7 = “completely happy“ (item 1) or 1 = “completely dissatisfied” to 7 = “completely satisfied“ (items 2 and 3), N = 1,099 (Germany), N = 33,725 (Total).

Objectivity

The scale was administered as face-to-face paper-and-pencil interview (PAPI) or computer-assisted personal interview (CAPI) in most countries. These interviews are personally conducted by specially instructed interviewers. Furthermore, the survey includes a standardized questionnaire format and written instructions. Fixed categories are used to rate the scale. As a result, a high objectivity of application is achieved. Furthermore, a high objectivity of interpretation is ensured, as a norming data is available for the scale (see descriptive statistics).

Reliability

The reliability of the scale score (mean of the three items) was estimated for each country separately. We estimated the reliability based on Cronbach’s alpha and Raykov’s ρ (Raykov, 1997). If items are tau-equivalent, Cronbach’s alpha is an unbiased estimator of the scale reliability. If items are tau-congeneric, Cronbach’s alpha is only a lower bound estimator of the reliability and Raykov’s ρ reveals a more accurate (and usually higher) reliability estimate. However, in the present sample, Raykov’s ρ was consistently lower than Cronbach’s alpha. We do not have an explanation for this unusual pattern of results and hope that future research will help to clarify this conundrum.

Table 4

Cronbach’s Alpha and Raykov’s Rho for each country and across all countries (total)

Note. Observations were weighted based on design weights, recoded scale from 1 = “Completely unhappy”/”Completely unsatisfied” to 7 = “Completely happy”/“Completely satisfied”.

Validity

Measuring the construct in a way that it is consistent with existing theories and definitions facilitate the construct validity. As all three items of the Happiness and Satisfaction Scale measure different aspects of happiness, the single items as well as the total score were correlated with other scales and variables from the ISSP 2012. It is assumed that a higher correlation of the total score with relevant variables reflects validity for an overall happiness value, in contrast to the single items reflecting specific aspects of happiness and satisfaction. First, there are numerous studies supposing a strong positive relation between health and happiness (e.g. Zajacova & Dowd, 2014; Touburg & Veenhoven, 2015). Therefore, the relationship between subjective health and the Happiness and Satisfaction Scale was examined in the ISSP data. It was expected that more healthy persons show a higher overall happiness and satisfaction score. Furthermore, according to Wu and Tam (2015) happiness can be related to socio-economic status (SES). There is reason to assume that people regarding themselves as more happy in life should perceive themselves to be rather at the top of society. Regarding satisfaction with job and family life, there should be a negative correlation with work-family conflict. Sometimes, job interferes with family life or the other way round, for example if there is not enough time to fulfil duties in both domains. If this is the case, the conflict between family and job should be related negatively with both satisfaction with job and family life. With respect to satisfaction with family life, there should be a positive correlation for persons with a functioning family life, which includes a steady relationship. This is supported by Halman (1987), who reported a positive effect of being married as an indicator for a steady relationship on happiness and satisfaction. Furthermore, if having children and living together with one’s family is more relevant to someone, there is a higher probability of living together with a higher number of persons in the household. Therefore, a higher number of persons in the household can be positively correlated with more satisfaction with family life. The same applies to a positive attitude towards children, because considering children as a positive contribution to one’s life should lead to a higher satisfaction with family life.

Finally, according to several studies analyzing the relationship between income and happiness (e.g. Noelle-Neumann, 1977; Halman, 1987; Cantril, 1965; Neuberger & Allerbeck, 1978), there should be a positive correlation between higher income and happiness, especially with satisfaction with job.

Therefore, the following scales from ISSP 2012 were used to examine construct validity:

-       Health:

-       “In general, would you say your health is excellent/very good/good/fair/poor?” (-)

-       Top-Bottom self-placement:

-       “In our society, there are groups which tend to be towards the top and groups which tend to be towards the bottom. Below is a scale that runs from the top to the bottom. Where would you put yourself on this scale?” – scale from 1 (bottom) to 10 (top)

-       Work-family conflict (α = .77): “How often has each of the following happened to you during the past three months?”

-       “I have come home from work too tired to do the chores which need to be done.”

-       “It has been difficult for me to fulfil my family responsibilities because of the amount of time I spent on my job.”

-       “I have arrived at work too tired to function well because of the household work I had done.”

-       “I have found it difficult to concentrate at work because of my family responsibilities.”

-       Number of persons in household:

-       “Including yourself, how many people - including children - usually live in your household?” – scale from “1 person” to “5 or more persons”

-       Living in steady partnership:

-       “Are you living in a steady partnership?” (-) – “1. Yes, have partner; live in same household, 2. Yes, have partner; don’t live in same household, 3. No partner”

-       Attitude towards children (α = .50): “To what extent do you agree or disagree"?”

-       “Watching children grow up is life's greatest joy.”

-       “Having children interferes too much with the freedom of parents.” (-)

-       “Children are a financial burden on their parents.” (-)

-       “Having children restricts the employment and career chances of one or both parents.” (-)

-       “Having children increases people’s social standing in society.” (-)

-       “Adult children are an important source of help for elderly parents.”

-       Country specific personal income:

-       “Before taxes and other deductions, what on average is your own total monthly income?”

Table 5

Correlations of the Happiness and Satisfaction Scale with relevant variables

Note. Pearson correlation coefficients, N = 53,163. All correlations are significant, *p < .05, **p < .001 (two-tailed). Observations were weighted based on design weights. The following variables were not asked in every country: self-placement in society (United States), number of persons in household (Turkey), living in steady membership (Denmark, Great Britain, Taiwan), Attitude towards children (Spain).

Practical importance of the validity coefficients can be interpreted according to Cohen’s (1992) standards: small effect (r = .10), moderate effect (r = .30), strong effect (r = .50). Consistent to expectations, overall happiness and satisfaction correlates positively at a small to moderate level with health and self-placement at the top of society. There is also a small negative correlation of overall happiness and satisfaction, including all three aspects (happiness in life, satisfaction with job and satisfaction with family life), with work-family conflict. Furthermore, there is a small positive correlation between satisfaction with family life and living in a steady relationship as well as with the number of persons in the household and a positive attitude towards children.

Regarding satisfaction with job, there are small to moderate correlations with personal income, especially in countries with lower gross domestic product (GDP). The results can be explained by studies suggesting a positive correlation of income and happiness only until basic needs are fulfilled (e.g. Di Tella & MacCulloch, 2010; Wang, Pan, & Luo, 2015). In order to better illustrate the differentiated relation between satisfaction with job and income, correlations are shown as average scores across all countries (total) as well as for every country separately in Table 6.

Altogether, the expected relations could be confirmed by data. The results indicate validity for the Happiness and Satisfaction Scale.

Table 6

Correlations of the Happiness and Satisfaction Scale with country specific income

Note. Pearson correlation coefficients. *p < .01, **p < .001 (two-tailed). Observations were weighted based on design weights. In Turkey, country specific income was not asked.

Descriptive statistics (scaling)

Table 7 presents reference means and standard deviation norms for the Happiness and Satisfaction Scale for the single items as well as for the mean score for all countries. The values for skewness range from -1.15 to -.01 (United States). The values for kurtosis range between -.46 (Venezuela) and 3.34 (Mexico). According to West, Finch and Curran (1995), deviations from normality for │skewness│ ≤ 2 and for │kurtosis│ ≤ 7 can be neglected. Therefore, the assumption of a normal distribution can be accepted.

Table 7

Reference values for each country and across all countries (total)

Note. Observations were weighted based on design weights, recoded scale from 1 = “Completely unhappy”/”Completely unsatisfied” to 7 = “Completely happy”/“Completely satisfied”.

Further quality criteria

To test fairness between countries for the Happiness and Satisfaction Scale, measurement invariance was investigated by structural equation modeling. We investigated three possible levels of measurement invariance: When configural invariance is given, the factor structure is equivalent between countries. Metric invariance implies that the factor structure and the item loadings are equivalent. For scalar invariance, the factor structure, the item loadings and the intercepts need to be equivalent across countries. As the configural model for this scale is perfectly identified and can therefore not be tested, only metric invariance and scalar invariance were investigated. To test metric invariance, all items were constrained to load on the same factors and the factor loadings were constrained to be invariant across countries, whereas the variance of the latent happiness and satisfaction variable and the items’ intercepts were allowed to differ between countries (see Figure 1). To investigate scalar invariance, all items were constrained to load on the same factors and the factor loadings as well as the items’ intercepts were constrained to be invariant across countries. As suggested by Chen (2007), we used the change in CFI as decision criteria whereby a difference of ΔRMSEA ≤ .015 and a ΔCFI ≤ .010 between the configural and the metric invariance model suggests metric invariance; and a difference of ΔRMSEA ≤ .015 and a ΔCFI ≤ .010 between the metric and the scalar invariance model suggests scalar invariance. Results are shown in Table 8 below.

Table 8

Measurement invariance for Happiness and Satisfaction across all countries

Note. Observations were weighted based on design weights, N = 53,532, ***p ≤ .001.

The model of metric invariance shows a good fit across countries. However, for the model of scalar invariance, CFI and RMSEA suggest a low model fit. This is confirmed by a comparison of the chi-square-values, which reveals a big difference between the two models (ΔCFI = .15). According to Chen (2007), only the model of metric invariance can be accepted.

The results suggest equivalent meanings of the items of the Happiness and Satisfaction scale, but different item intercepts (item difficulties) across countries.