SUMMARY
In contrast to cross-sectional research, time-series designs generally challenge the researcher's ability to discover and analyze multiple-item, multiple-method data required for reliability and validity assessment. Cur
rently archived time-series data generally include at best single-item measures infrequently available for a restricted period of time. Increased attention to measurement issues in time-series research should result in renewed emphasis on the design, measurement, and archiving of multiple "focal local indicators" (Campbell 1976) of marketing phenomena. It also should contribute to further research in the limitations of currently available time-series statistical models and causality tests and to the development of promising altemative procedures (see Snetsinger 1983 for a partial summary