Test-retest reliability of the Tower of London Planning Task (TOL-F)

Test-retest reliability is difficult to establish for measures of executive functioning that rely on task novelty. Correspondingly, evidence on the test-retest reliability of the commonly used Tower of London (TOL) planning task is, as yet, equivocal and only based on indices of relative consistency, rather than absolute agreement of individual scores. Further, the stability of planning latencies over repeated testing has not been investigated. The present study assessed test-retest reliability of planning performance measures using a structurally balanced problem set implemented in the TOL-Freiburg version (TOL-F). The TOL-F was administered in 2 structurally identical versions to a sample of young, healthy adults over a 1-week interval. For planning accuracy, the Pearson correlation and intraclass correlation coefficient for relative consistency were adequate (r = .739 and .734), with the intraclass correlation coefficient for absolute agreement only slightly decreased (r = .690). For initial thinking and movement execution times, relative consistency and absolute agreement reliability indices were uniformly low (all r between .274 and .519). Given adequate planning accuracy test-retest reliability, the TOL-F can be reliably used to measure planning ability in group-based studies and with individual participants, as is important for clinical testing. Planning latencies, however, should only be used as complementary, but not sole measures of planning ability, particularly for normative evaluations in clinical assessment. In sum, TOL-F planning accuracy possesses adequate absolute and relative test-retest reliability for experimental utility. Future studies should assess whether this indeed translates into clinical utility of the TOL-F for measuring planning ability in patients.