1 Introduction

The concept of generativity, which was introduced to psychology by Erik Erikson, was later elaborated by many authors. According to Erikson, generativity is primarily “the concern in establishing and guiding the next generation” (Erikson 1984, p. 240).

One of the widest elaborations of generativity today is the theoretical model proposed by McAdams and de St. Aubin (1992). These authors explain generativity as a transitional construct which encompasses individual, interpersonal and social dynamics. Consequently, their model of generativity includes seven interrelated components: cultural demands, inner desire, generative concern, commitment, belief in the species, action and generative narration.

This model has not yet been tested completely, that is only some of its components and their relations have been tested. The main reasons for such an empirical status of the model are inadequately clear operationalizations of the components of the model. The best components by the authors of the model and their associates are generative concern and generative action. They constructed the well-known Loyola Generativity Scale that measures generative concern and the Generative Behavior Check List. The component of belief in the species is described relatively well, but for research purposes is not operationalized adequately. For the purpose of research we operationalized this component in terms of Wrightsman’s Philosophies of Human Nature Scale (1974) (Wrightsman 1992). We have validated and adapted this scale on samples of adults from Croatia and Slovenia under the title of “Philosophies About Man” (Ćubela Adorić et al. 2005a, b). The scale measures two dimensions: trust in people and cynism. Previously, we adapted the Loyola Generativity Scale (Lacković-Grgin et al. 2002; Lacković-Grgin 2004; Poljšak-Škraban and Žorga 2005). In order to test relations of other components of the model of generativity, we adapted the Generative Behavior Check List and applied it to Croatian and Slovenian adult samples (Tucak et al. 2005; Tucak 2006).

Generative action as a behavioral component of generativity is under the strong influence of cultural demands (i.e. socio-cultural forces) and inner desires. Every culture expects some form of responsibility from adults toward younger generations, the creation of new material and cultural norms as well as the preservation of past ones. Culture influences the forms of expression of generative aspirations and generative actions. In modern Western societies ideas like freedom, growth, development and progress of the individual and society as a whole, are very appealing. At the end of the 20th century those ideas were played a role in the political and social changes in eastern European countries. A better future was promised to the people of these, nowadays, transitional countries. However, a promise remains attractive only if the object of aspiration is not deranged and destroyed considerably, that is, when progress, not regression, is visible as compared to earlier states. Different transitional countries have not achieved the same degree of expected progress. There are a number of social and economic indicators showing differences in the effects of transition between two of the most developed republics of former Yugoslavia, Slovenia and Croatia. For example, in the period from 1993 to 2001 Slovenia reached the highest mean rate of growth among all post communist countries which have joined the European Union (by May 1, 2004), while at the same time Croatia had not reached the level of development which it had before the beginning of the transition (Šušteršić 2004). At the beginning of the 21st century Slovenia had 130% higher GDP than Croatia (Mihaljek 2003). In short, Slovenia experienced greater progress in this time period than had Croatia. Taking these factors into consideration we organized a comparative study. Our assumption was that described differences could affect some aspects of generativity, such as the inner desire to be generative or the evaluation of the importance of some life areas in which generative action is expected, since adults have many more or less generative roles in these areas.

The inner desire to be generative does not always correspond to real circumstances. It is since people could want something regardless of the possibilities of achieving that. Therefore we would like to suggest that goals are more important in the prediction of generative action than inner desire. Goals are the result of internalized cultural demands and norms, as well as inner, biological forces. They organize complex patterns of behavior, thereby generative action as well. As a result we opted to operationalize inner desire, the component of the McAdams and de St. Aubin’s model, in terms of life goals. For this purpose we used GOALS: A Questionnaire for measurement of life goals (Pöhlmann and Brunstein 1997). The rationale for choosing this instrument is the fact that, while designing the questionnaire, the authors rely on Bakan’s well-known theorizing about two basic human orientations: agency and communion (Bakan 1966). The authors have applied the questionnaire to German students. On the basis of factor analyses of questionnaire data, the authors found that it measures life goals that are relevant for generative action. These include the goals of power, achievement and diversity (agency), and the goals of intimacy, altruism and affiliation (communion).

Earlier research has shown that various expressions of generativity are linked to motives for activity and for community (Peterson and Stewart 1993, 1996). Generativity, whether treated as a homogeneous or as a multi-faceted concept, was in earlier studies brought in relation with different variables of psychological well-being (de St. Aubin and McAdams 1995; Grossbaum and Bates 2002; Peterson and Klohnen 1995; Tucak 2005), parenthood, positive practices of upbringing and authoritative style parenthood (Peterson et al. 1997; Snarey and Clark 1998; Van Hiel et al. 2006), personal characteristics such as extraversion, pleasantness, openness to new experiences, pro-social characteristics (de St. Aubin and McAdams 1995; Peterson and Klohnen 1995), etc. While caring for ones own children and family represents a private and narrower area of expression of generativity, the inclusion of social and political activities presents the possibility of expressing generativity on a wider and on a public level. In this context there are a series of studies that have shown that generativity is linked to greater social and political engagement, such as, for example, volunteering in charitable organizations, fighting for civil rights, voting, etc. (Peterson and Klohnen 1995; Peterson and Stewart 1996; Peterson and Duncan 1999).

In this study we have tried to determine: first, if there are any differences in the evaluation of importance of certain life areas (employment, marital status, parenthood and free time), life goals, generative concern and generative action between adults from the two transitional countries, and second, how important are the evaluations with regard to the mentioned life areas, life goals, and generative concern in the prediction of generative action in Croatian and Slovenian adults.

2 Method

2.1 Participants/Subjects

During the year 2004 this study included 381 adults from Croatia (aged 21–67). Participants were from different regions of Croatia (Western Croatia, Slavonia, Dalmatia, Istria, Lika and Gorski Kotar). Forty-three percent of participants were male and 57% were female. This sample was heterogeneous not only regarding basic demographic variables (age and gender), but also educational level and profession, as well.

At the same time, in different regions of Slovenia 546 subjects (aged 24–70) participated in the study. The sample included 46% of male and 54% of female participants of different levels of education and heterogeneous professions.

In key sociodemographic variables, significant differences between the two samples of participants have not been established—in gender (χ2 = 1.28, P = .258), employment (χ2 = 2.34, P = .126) and age (t = 1.519, df = 924, P = .128).

2.1.1 Procedure

Prior to undertaking the examinations, each examiner was instructed on how to undertake the examination. The instruction consisted of a short description according to which each examiner was to find participants that according to their sociodemographic characteristics (sex, age, employment status) met the requirements of the examination. Before filling out the questionnaire, the participant was informed of the principal aim of the examination. The examination was done individually, meaning the examiners handed the questionnaire separately to each participant in a sealed envelope, either at the participant’s home or at his/her place of employment. According to the arrangement with the participant, the examiner would either wait on the spot until the participant filled in the questionnaire, or else would pick up the questionnaire the following day.

2.2 Measures/Instruments

The same questionnaire was applied to both, Croatian and Slovenian samples. The questionnaire included questions about sociodemographic data (age, gender, years of education, professional status, employment status, marital status and number of children). There were also questions regarding the importance of employment status, the importance of working in the profession for which they finished college/school, the importance of marital status, parenthood and the importance of free time.

Participants assessed the importance of these different fields on a scale of five degrees (from 1—not important at all, to 5—very important). Furthermore, this questionnaire included several other scales. For the purpose of this study, the data of the following scales were analyzed:

2.2.1 Loyola Generativity Scale-LGS, Mc Adams and de St. Aubin (1992)

The original scale contains 20 items which assess generative concern as the main component of McAdams and de St. Aubin’s model of generativity. The items tap into many of the main content domains of generativity. Examples of typical items include: I try to pass along the knowledge I have gained through my experiences, Other people say that I am a very productive person, People come to me for advice, etc. The authors of this scale reported that reliability in the adult sample is quite high (Cronbach alpha = .83), and that convergent validity is satisfactory. These characteristics of the scale were confirmed in our earlier studies (Lacković-Grgin et al. 2002). Very small differences in number of valid items were found between these samples after factor analysis was done, separately for the Croatian and Slovenian sample. For the purpose of comparing results of the two samples it was decided that factor analysis should be conducted on the combined data of both samples. Based on this analysis, a unitary version of LGS was created. This scale included 17 items and the Cronbach alpha was .84. Compared with the original scale, which was created in American culture, in our adapted version of LGS some items were dropped out. Those items referred to voluntary work for a charity, caring for the homeless and adopting children. In this study we used data of 17 items of the LGS version.

2.2.2 Generative Action Scale

Generative Action Scale is a result of an adaptation of the Generative Behavior Checklist (McAdams and de St. Aubin 1992). Originally, the scale contains 50 items, 40 of which assess generative behaviors while 10 items refer to behaviors irrelevant for generativity. In the original checklist the participant had to mark the performing frequency of described behavior during the past 2 months (0—never, 1—only once, 2—more than once). In our study, the original checklist was adapted into the scales with instructions for the participant to give an assessment on a five degree scale (from 1—never or very rarely, to 5—very often or almost always) how frequently he/she performed the described behavior in the past 2 months. It seems that an adapted version allows for a better examination of individual differences regarding the level of generative action than the original checklist.

Further adaptation related to changing some items from the original checklist. These changes included keeping 11 original items while nine new items were added. Since the checklist originated from North-American culture, we added some new items that are more specific for the cultures in question. Researchers in this study which were formerly introduced to McAdams and de St. Aubin’s generativity model and to the concept of generative action generated new items. After the psychometric validation of an adapted scale, generative action in our study was assessed with this 20 items scale. Typical items in this scale are: Taught somebody a skill, Did something that other people considered to be unique and important, etc. In a preliminary sample of students, internal consistency of the whole scale was .88 (Cronbach alpha), while in this sample Cronbach alpha was .87.

It is important to notice that results of common factor analysis showed the existence of two factors. The first factor we have labeled as Specific generative action (e.g. Restored or renovated a house, part of a house, a piece of furniture etc.). Regarding the content of the items, we have labeled the second factor as General generative action (e.g. Made a decision that influenced many people). Since these two factors correlated significantly, in addition to other psychometric arguments, in this study we used this 20 items scale as a measure of overall generative action. (More information about the scale and its psychometric characteristics can be found in Tucak (2006)).

2.2.3 Life Goals Questionnaire

Life Goals Questionnaire which we have used in this study is a shorter, 16 items version of the original 24 items questionnaire (Pöhlman and Brunstein 1997). Eight items measure power and achievement, e.g. agency and the other eight items assess altruism and intimacy, e.g. communion. Exploratory factor analysis confirmed two factors (two orientations), that we have named agentic goals and communal goals. Items that have loadings on agentic factor are: e.g. Be on prestigious position, Develop own skills, etc. Typical items for the communal factor are: e.g. Do something good to others, Give affection and love, etc. Reliability coefficients of subscales are satisfactory for research purposes, Cronbach alpha are 0.72 for agency, and 0.80 for communion.

3 Results and Discussion

First, we have computed descriptive parameters for both samples. Due to the fact that some subjects did not answer all the items of scales, the number of subjects have varied through variables, as can be seen in Table 1.

Table 1 Basic descriptive parameters of measured variables in Croatian (N = 381) and Slovenian (N = 546) sample
Full size table

In order to test any differences between Croatian and Slovenian subjects in measured variables, t-tests were computed, and the results are shown in Table 2.

Table 2 Results of t-tests for comparison of measured variables between Croatian and Slovenian adults
Full size table

From Table 2, is evident that Croatian and Slovenian subjects differ predominantly in generative concern, but they also differ significantly in evaluation of importance of employment, importance of free time, and in agentic and communal goals. There are no significant differences in other variables between the two samples.

Employment is more important for Croatian than for Slovenian adults. This difference can be explained by different rates of unemployment in the two transitional countries. Since the beginning of the 1990s there has been a high rate of unemployment in Croatian society, not only among young people, but also among older adults who lost their jobs because of economic re-organization. Having a job is a basic existential need for most people, so it is not surprising that adults in both countries find employment very important. Based on international studies of values, conducted in certain European countries, it can be concluded that the importance of work decreased in Slovenia from 73 to 62% in the period from 1990 til 1999 (Rus and Toš 2005). Today, however, work is more important as expression of personal growth for Slovenian people than as an expression of traditional moral standing that was more dominant previously. The cited authors believe that this is an indication that Slovenian people are beginning to resemble people in Scandinavian countries with regard to their work values.

The importance of free time is significantly higher among Slovenian subjects and that could be the result of postmodern processes which are, among other things, characterized by a decrease in the value of work and an increase in the value of free time. Rus and Toš (2005) state that the value of free time in Slovenia has increased from 28 to 33%.

Croatian and Slovenian adults also differ in agentic and communal goals, that is Croatian subjects attained significantly higher results on scales that measure these goals. If we were to interpret these goals as an inner desire to be generative, the difference could be the result of a higher level of achievement of these goals among Slovenian subjects. Therefore, it can be assumed that Slovenian subjects do not find them as salient as Croatian subjects who, most likely, expect to achieve them in a future, better time.

The greatest difference between the two samples is in reference to generative concern which is, theoretically speaking, one of the main components of generativity. Significantly higher levels of generative concern among Croatian adults can be interpreted as a consequence of the ravages of war as well as postwar criminal and corruptive actions. Due to these life circumstances many Croatian people are very anxious about their personal material and cultural standing in contrast to Slovenian citizens, and more anxious about the future of younger generations.

It’s interesting to consider correlations of measured variables among Croatian and Slovenian samples. They are shown in Table 3 and 4.

Table 3 Matrix of correlations of measured variables in the Croatian sample (N = 318)
Full size table
Table 4 Matrix of correlations of measured variables in the Slovenian sample (N = 450)
Full size table

Patterns of correlations are similar in the two samples. Significant, but slightly higher correlations between generative action and agentic goals, as well as communal goals, are found in the Slovenian sample (.32). In the Croatian sample these correlations are .21 and .27, respectively. This complies with our interpretation of differences in goals of agency and communion among Croatian and Slovenian subjects. These goals are evident in generative action of adults in Slovenia. The main difference is in the correlation of generative action and importance of free time. These variables correlated higher among Croatian adults (.33) than among Slovenian adults (.19). It seems that Croatian adults use generative action if they have more free time, for example, if they are not occupied with personal, existential problems. The results of our previous research of generative concern confirm this explanation (Lacković-Grgin et al. 2002). They showed that some items of the Loyola Generativity Scale that were valid in samples of American adults, are not adequate in Croatian culture. Those items referred to voluntary activities and adopting children. However, in an earlier study the same was shown to hold true for Slovenia. It was shown in the examination of samples of adult Slovenes that elements of the Loyola scale of generativity which apply to volunteering and adoption are no more valid in Slovenian culture (Tucak et al. 2005).

On the basis of the McAdams and de St. Aubin’s model, as well as on the basis of previous research of relation of generative action and generative concern, statistically significant, but expected, correlations have been obtained between these two constructs (.52 for Croatian adults, and .50 for Slovenian adults).

It is interesting to note that the relatively highest correlations are found between the importance of marital status and importance of being parent (.62 for Croatian adults, and .50 for Slovenian adults). A somewhat smaller correlation in the sample of Slovenian adults might be the consequence of greater modernity in Slovenian society in which even earlier, more than in Croatia, parenthood out of wedlock was tolerated.

The final analysis should answer the question of predictive value of importance of some life areas, life goals and generative concern for generative action of adults in Croatia and Slovenia. Results of stepwise regression analyses (method backward) are shown in Tables 5 and 6.

Table 5 Results of regression analysis with generative action as a dependent variable in the sample of Croatian subjects
Full size table
Table 6 Results of regression analysis with generative action as a dependent variable in the sample of Slovenian subjects
Full size table

As shown from Tables 3 and 4, correlations between generative action as dependent variable and predictive variables are significant and higher than between predictors. There is an exception in both samples: correlations between the importance of marital status and importance of being a parent. They should be taken into account because high correlation between predictors could have an impact on the accuracy of prediction. There are no significant differences in generative action but there are significant differences in predictors of generative action between two samples.

In the Croatian sample only three variables were significant in the prediction of generative action: the importance of free time, generative concern, and communal goals. These three predictors have explained about 34% of variance of generative action.

In the Slovenian sample significant predictors included: importance of being a parent, generative concern, agentic goals and communal goals. Although the number of significant predictors was higher in the Slovenian sample, they explained only a smaller percentage of variance of generative action (30%). However, predictors of generative action in the Slovenian sample are closer to the assumption derived from McAdams and de St. Aubin’s model of generativity, since they include both dimensions of goals, which authors defined as the internal desire to be generative. The importance of parenthood was significant in the prediction of generative action, but only in the sample of Slovenian adults. Previously it was shown that there are no significant differences between Croatians and Slovenian adults in the assessment of the importance of being a parent.

The aforementioned Slovenian sociologists state that family values in modern Slovenia have remained the same as before transition when experts propagated responsible parenthood (e.g. love and care for one’s children, a democratic style of upbringing), and not parenthood for its own sake (Toličić and Zorman 1977). In Croatia after the transition, parenthood for its own sake is being propagated from political and church circles. The influence of the church has increased, and unlike in Slovenia, the number of believers has grown from 39% in 1989 to 75% in 1996 (Skledar and Marinović-Jerolimov 1997). The results of our examination into the motivation for parenthood talk of the influence of such propaganda (Lacković-Grgin and Vitez 1997). While examining the categories of motivation for parenthood identified by Rabin (1965) and Rabin and Greene (1968), we have confirmed the existence of altruistic, fatalistic, and narcissistic motivation. Instrumental motivation consists of four factors: 1. Confirmation of self, 2. Extension of the family line, 3. Preservation of stability within a marriage and 4. Patriotic motivation. The fourth factor is not mentioned in other studies. Because of the war and emigration, those in Croatia want to have children “to increase the birthrate, to stimulate a demographic renewal, and to fulfill one’s national duty” (Lacković-Grgin and Vitez 1997, p. 55). The importance of parenthood motivated in such a way is not a good predictor of generative activity as it is presented in this study.

In the Croatian sample one of the significant predictors of generative action was the importance of free time. This means that people use their free time for generative action, and as has been stresses, goals of communion are important for that action. It is possible that Croatian adults work activities serve predominantly to satisfy their basic existential needs, while the higher order needs (e.g. need to belong, need to care and love) are satisfied in their free time. Although adults from Slovenia assess their free time as more important (Rus and Toš 2005), free time is not a significant predictor of generative action in Slovenian adults. This could indicate that in this sample, free time serves in the role of recreation and self-actualization. This function of free time could be found in more developed Western countries (Cutler and Hendricks 1990).

It should be noted that one of the dimensions of goals, goals of agency, which are related to power and achievement, have significant predictive value for generative action for adults from Slovenia only. Higher levels of social and economic development in Slovenia allow for greater possibilities of attaining those goals through generative action. This is less like by Croatian adults because Croatia has not yet reached that level of development. As mentioned in the description of instruments, the Generative Action Scale gives the possibility of calculating results for general and specific generative action. After comparing adults from the two countries with regard to their results for general and specific generative action, we found that they did not differ in specific generative action, but Slovenian adults scored significantly higher on the general generative action subscale. This is in accordance with the assumption that more favorable social circumstances have a positive impact on general generative action.

4 Conclusion

Our assumptions about the impact of different levels of economic and social development on generative action of adults from two former Yugoslavian republics were partially confirmed. Although there were no differences in total generative action between adults from Croatia and Slovenia, there are differences in importance in certain life areas, in generative concern, and in life goals as predictors of generative action. There are also differences in the number and type of significant predictors of generative action. Although there are significant differences in generative concern among Croatian and Slovenian subjects, that key component of generativity has the greatest predictive role for generative action in both, the Croatian and Slovenian sample. This is in accordance with other theoretical viewpoints, as well as empirical data.