Immediacy and Learning Meta-Analysis 17
22,915, 27% shared variance) and those of experimental studies (average r = .306, k = 7, N =
1,559, 9% shared variance). These experiments were constrained by factors such as time of
teacher immediacy displayed, specific teachers, content-topics of messages, the specific
situation, the brevity of the inductions, etc. Despite accounting for 18% more variance and the all
too frequent use of causal language, the surveys did not, and could not, establish the causal link
between teacher immediacy and student learning. These results may be inflated by uncontrolled
variables associated with immediate teachers. Given the limitations of both methods, the true
estimate of the association between teacher immediacy and student learning, while elusive, may
lie between the estimates provided by experiments and surveys.
These analyses of overall learning and overall teacher immediacy provided general
conclusions that facilitated a broad perspective on the relationship of teacher immediacy to
learning. However, the testing of variances within all the groupings discussed above revealed
greater variance than would be expected from random sampling error alone. Results from
heterogeneous samples may be useful to affirm the direction and relative strength of associations,
but we continued our analysis by examining three specific categories of immediacy and three
specific categories of learning. This further testing provided more analytic insights and specific
conclusions.
Three Types of Immediacy and Three Types of Learning
Drawing from Bloom’s (1956) taxonomy of educational objectives and general tenets of
attitude theory
(Inskso, 1967; Triandis, 1971; Zimbardo & Ebbesen, 1969),
we classified student
learning outcomes into three categories: perceived learning, cognitive learning, and affective
learning. Because of the large number of studies using the learning loss measure, we created the
learning category of perceived learning. Although we believed that learning loss measurement
Convention