The first task in assessing longitudinal trends is to decide on an appropriate measure of creativity. For the most part, researchers have adopted one of two assessment methods: psychometric tests and productivity indicators.
Psychometric tests. A large number of psychometric instruments exist that purport to assess creativity. Some measures assess personality characteristics, others cognitive style, and yet others biographical background factors. Nonetheless, research on the relation between age and creativity has been almost exclusively confined to one particular set of tests, namely, those that purport to assess a person's capacity for "divergent thinking." Such measures determine whether an individual can generate an impressive number of novel responses to test stimuli. Typical is the unusual uses test that requires the respondent to conceive all of the potential uses for a paper clip. These tests can be scored for fluency (number of responses), flexibility (number of distinct categories to which the responses belong), and originality (how rare the response is relative to others taking the test). Moreover, some divergent-thinking tests use verbal stimuli, whereas other tests use visual stimuli. The underlying assumption behind these measures is that they tap the cognitive processes that are essential to creative thought. In any case, investigators who have applied such divergent-thinking tests have consistently found that divergent thinking tends to exhibit an inverted-U shape with regard to age over the life course, with a clear tendency for scores to drop off in the latter half of life. Optimum creativity usually appears around the fortieth year of life.
Even so, caution must be exercised in interpreting these findings. It cannot be confidently inferred from these results that creativity must decline after a person attains middle age. First of all, most of these empirical investigations depend exclusively on cross-sectional data, a methodological tactic that conflates the effects of aging with those of birth year (i.e., age versus cohort effects). Hence, special care must be taken to gauge aging effects from truly longitudinal data (within, rather than across, individuals). In addition, the specific shape of the longitudinal trajectory is contingent on the particular types of tests that are used. Because divergent-thinking tests constitute only one possible type of creativity assessment, different age curves can emerge when different measures are applied. Most tellingly, instruments that assess problem-solving abilities in more practical situations can actually yield scores that fail to decline with age, and may even increase. Lastly, not all experts in the area of creativity assessment accept the validity of so-called creativity tests. Most validation studies reveal that such tests exhibit small correlations with direct behavioral measures of creativity, such as achieved eminence in a creative domain.
Productivity indicators. The best single predictor of achieved eminence as a creator is lifetime creative output. Therefore, it appears reasonable to adopt productivity as a behavioral measure of creativity. Because scientific inquiries into age changes in creative productivity began in 1835, with the work of Adolphe Quételet (1796–1874), this topic can be considered the oldest in life-span developmental psychology. Yet the first truly important figure in this area was Harvey C. Lehman, whose work was summarized in his 1953 book Age and Achievement. Although Lehman's research was plagued with many methodological problems, later investigations that introduced more advanced methods have confirmed his central conclusion: the generation of products tends to increase with age until a maximum output level is attained, with productivity declining thereafter. Indeed, the age where output peaks corresponds approximately with the age where performance on divergent-thinking tests usually maximizes.
Nevertheless, research using productivity measures also have positive implications for the expected level of creativity in the later years of life. This optimism follows from seven empirical results:
- The age-curve specific form—especially the placement of the peak and the slope of the postpeak decline—depends on the domain in which creativity takes place. In some domains the optimum will occur much later in life, and the drop will be very slow, or even negligible.
- Creative productivity seldom declines to zero. On the contrary, in most creative domains, persons in their seventies will display higher output rates than they did in their twenties. Furthermore, those in their seventies will usually be generating ideas at a rate only 50 percent below what they achieved during their productive peaks.
- Individual differences in lifetime productivity are far more substantial than longitudinal changes in productivity within any particular creator's career. In other words, cross-sectional variation in output accounts for more variance than does age. Accordingly, highly prolific creators in their seventies and eighties are more productive than are less prolific creators during their career acme.
- Longitudinal fluctuations in creative output are a function of career age, not chronological age. Hence, "late bloomers" who begin their careers much later in life will not reach their career optima until much later in life. The same pattern holds for those who switch fields, thereby resetting the longitudinal clock.
- A respectable amount of the productivity loss in the last half of life is not necessarily inevitable, insofar as it can be ascribed to various extrinsic factors, such as declining health or increased professional or personal responsibilities. By the same token, certain settings can sustain creativity well into the later years. In the sciences, for instance, those creators who are enmeshed in a rich network of colleagues and students are prone to exhibit longer creative careers.
- If one looks at the quality ratio of successful works relative to total works produced in consecutive age periods, one discovers that this ratio does not change systematically over the course of a creative career. Most notably, this success rate does not diminish as a creator ages. As a result, although creative elders may produce fewer masterpieces in their final years, they also generate fewer inferior works. On a work-for-work basis, there is absolutely no reason to speak of any age-related decrement.
- Quantitative declines in creative productivity across the life span are often accompanied by qualitative changes in the nature of the output—changes that frequently operate in a compensatory manner. For instance, as creators mature, they will tend to focus on more ambitious products, such as epics, operas, novels, and monographs. More critically, creators in their concluding years often greatly alter their approach to their creative endeavors. In the visual arts, this longitudinal shift is called the old-age style, while in music this change is styled the swan-song phenomenon.
The foregoing considerations imply that psychometric measures may underestimate the creativity of older persons.