Factor structure and invariance of personal meaning measures in cohorts of younger and older adults

Abstract

The purpose of the present study was to examine the factor structure and factorial invariance of measures of personal meaning (i.e. the existential belief that life has purpose and coherence) in cohorts of younger and older adults. Samples of 163 young adults and 144 older adults completed six measures of personal meaning in life. Confirmatory factor analysis and multiple-groups confirmatory factor analysis was used to assess structure and structural invariance. The results show that a general second-order model of personal meaning best characterizes the factor structure of personal meaning of both younger and older adults. Factorial invariance across younger and older adults was demonstrated at the first-order factor level, but not at the second-order level. A test of age differences between the means of the six latent personal meaning factors revealed no significant differences, although there was a tendency for older adults to experience greater meaning in life. Implications for the use of personal meaning measures in cross-sectional and longitudinal research are discussed.

Introduction

Given the recent resurgence of interest in the field of positive psychology (Seligman & Csikszentmihalyi, 2000, Taylor et al., 2000), there has been increasing research activity in the study of personal meaning across the life span (Reker & Chamberlain, 2000, Wong & Fry, 1998). During the past three decades, a relatively large number of empirical studies have clearly demonstrated that meaning in life is an important variable in the maintenance and enhancement of physical (e.g. Reker, 1994, Reker & Butler, 1990), psychological (e.g. Newcomb & Harlow, 1986, Nicholson et al., 1994, Phillips, 1980, Reker et al., 1987, Zika & Chamberlain, 1992), and mental (e.g. Fry, 2001, Harlow et al., 1986, Pearson & Sheffield, 1989, Reed, 1991, Reker, 1997) health. Many of these studies used different operational measures of meaning, examined the correlates of meaning, described the moderating and mediating role of meaning, or focussed on age or sex differences in meaning.

One of the difficulties with the concept of personal meaning is the proliferation of theoretical and conceptual approaches that make it somewhat problematic to integrate the empirical findings. While traditionally a philosophical/theological concept, it was the logotherapist Frankl (1963) who first introduced the construct of meaning in life to the social sciences. Frankl (1963) asserted that the will to meaning is a significant and universal human motive, the loss of which is characterized by boredom, hopelessness, depression, and the loss of the will to live. Yalom (1980) broadened that perspective by focussing on the ultimate human conditions of existence and using them as guidelines for existential psychotherapy. Maddi (1970) described the failure in finding meaning in life in terms of existential sickness or neurosis, the extreme form of which has the cognitive, affective, and conative properties of meaninglessness, apathy, and aimlessness. Reker and Wong (1988) incorporated the views of Frankl and Maddi and conceptualized personal meaning as a multidimensional construct consisting of the cognizance of order, coherence, and purpose in one's existence, the pursuit and attainment of worthwhile goals, and the accompanying sense of fulfilment. Thompson and Janigian (1988) focussed on understanding the search for meaning and offered the conceptual framework of life schemes that provide a sense of order and purpose to one's life. Baumeister (1991) conceptualized four different needs for meaning including the need for purpose, value, efficacy, and self-worth. These four needs for meaning offer a framework for understanding how people make sense of their lives. When these needs go unsatisfied, the individual shows signs of frustration, malaise, discontent, or instability.

A second related source of difficulty with the construct of personal meaning exists at the measurement level. Diversity in conceptualization is usually accompanied by a variety of measurement tools, each reflecting the theoretical orientation of the researcher. To date, a number of quantitative and qualitative instruments have been developed that can be categorized under three headings: general measures of meaning in life, domain-specific measures, and context-specific measures of meaning (Reker, 2000). Context-specific measures assess the meaning within the context of a specific experience, such as life-threatening illnesses (e.g. AIDS, cancer, heart attack). These are largely measures of how meaning is constructed, searched for, or found in the specific experience. Domain-specific measures assess the extent to which specific activities, goals, relationships, or values contribute to a meaningful life (e.g. personal relationships, creative activities, helping others). Domain-specific measures identify the sources from which an individual derives a sense of meaning and purpose. General measures of meaning reflect an individual's understanding of how events in life fit into a larger context as facilitated by a sense of coherence and a sense of purpose. As such, general measures assess an individual's existential belief system. Of the three categories of meaning measures, only the quantitative self-report measures of personal meaning are of primary interest in the present study.

When different measures of personal meaning are used across several studies, it is not known whether the scales, as a collective, measure the same underlying construct. This places a constraint on researchers who wish to conduct meta-analytic investigations of meaning in life. In order to feel confident that the different scales are measuring the same underlying construct, it is necessary to examine the factor structure of a number of available measures of personal meaning in life. Only one prior study addressed this issue. Chamberlain and Zika (1988) examined three scales in a sample of 194 women (average age 29 years): the Purpose in Life (PIL: Crumbaugh & Maholick, 1969) test, the Life Regard Index (LRI: Battista & Almond, 1973), and the Sense of Coherence (SOC: Antonovsky, 1987) scale. The scales were assessed using principal axis factor analysis with oblique rotation. At the first-order item level, all three scales were found to be multidimensional: four factors for the PIL, six for the LRI, and seven for the SOC. However, at the second-order scale level, a single higher-order factor emerged, suggesting that a ‘general meaning in life’ dimension underlies these measures. Thus, the Chamberlain and Zika (1988) study clearly demonstrates that core aspects of personal meaning, including purpose in life, framework, fulfilment, and sense of coherence, converge to form a unified underlying construct. The present study extends these findings by assessing the factor structure of additional scales of personal meaning that measure purpose, coherence, and self-transcendence.

Life span researchers are mostly interested in studying age differences and age changes in levels of meaning in life. Frankl (1963) theorized that meaning in life must be discovered over time. Reker and Wong (1988) postulated that the personal meaning system becomes more integrated with age. Indeed, Dittmann-Kohli and Westerhof (2000) provide empirical evidence that the personal meaning system becomes more integrated and consolidated with increasing age. Reker et al. (1987) found that older individuals experience a stronger sense of purpose in life, while younger adults show a stronger desire to achieve new goals and to anticipate a more meaningful future. Older adults have also been shown to derive greater meaning from involvement in religious activities, social causes, altruism, maintaining traditions and culture, and preserving cultural values and ideals. On the other hand, younger adults derive more meaning from meeting basic needs and personal achievements (Reker, 1988).

Taken together, these findings suggest that there are interindividual differences between younger and older adults in the experience of personal meaning. The equivalence of measuring instruments is often assumed in multigroup research and is rarely, if ever, tested directly. Thus, in addition to demonstrating that personal meaning is a coherent construct, it is also important to establish that the measurement and structure of meaning remains invariant across different age cohorts and over time. Demonstration of measurement and structural invariance would allow researchers to conduct cross-sectional and longitudinal studies without having to worry about whether observed differences and/or changes in meaning in life are due to age-related factors or to structural change on the meaning construct. However, when measurement and structural invariance cannot be demonstrated, observed differences may be confounded by other factors including differences in the way meaning is operationalized, measurement error, or instability in the meaning construct itself.

While it is important to demonstrate both cross-sectional and longitudinal invariance of personal meaning, design and time constraints did not allow us to collect data at more than one point in time. Therefore, this study will only assess the invariance of personal meaning across two age cohorts; namely, younger and older adults. Confirmatory factor analysis and multiple-groups confirmatory factor analysis were used to examine the factor structure and to test for factorial invariance, using EQS methodology (Byrne, 1994). Multiple indices were used to evaluate goodness of fit.

Prior to determining that the underlying meaning construct remains age invariant, it is necessary to examine the factor structure of personal meaning measures separately in each age cohort. Based on previous research, we propose that personal meaning scores are most adequately represented by a hierarchical factorial structure in which the covariation among the first-order meaning measures are explained by a single higher-order factor of general meaning, labelled existential meaning. As a general, over-arching construct, existential meaning encompasses the processes of discovering, creating, and searching for meaning including the core aspects of purpose in life, sense of coherence, life-goals framework, fulfilment of life goals, and self-transcendence (Reker & Chamberlain, 2000). As a test of the proposed factor structure, we hypothesized a second-order confirmatory factor analysis (CFA) model, as depicted in Fig. 1. In order to obtain appropriate indicators of each personal meaning measure, respective items were randomly divided into three sets of item parcels. The item parcels served as indicators for five of the six first-order personal meaning scales. For the multidimensional Life Regard Index, the subscales of Framework and Fulfilment served as the indicators. Each item parcel/subscale was hypothesized to have a nonzero loading on the first-order factor it was designed to measure and zero loadings on the remaining first-order factors. Error terms associated with each item parcel were hypothesized to be uncorrelated.

The main purpose of this study was to examine the underlying factor structure and the factorial invariance of personal meaning measures in cohorts of younger and older adults. It is hypothesized that a second-order factor structure will account for most of the variance in the personal meaning measures for each age cohort. Furthermore, it is expected that the first-order and second-order factor loadings will remain invariant across the two age groups. Finally, based on previous findings reported in the literature, we expected the older adults to experience greater personal meaning in life compared to their younger counterparts.