ElmCore Journal of Educational Psychology October, 2014 | Page 34

Science-Fellows®
establish the degree of error and inconsistency in
scoring, marking, grading, or evaluation, while the
former group proceed as if it is irrelevant. As we have
argued elsewhere: Humans can introduce error into
scores of performances (e.g. essays, portfolios, and
dramatic or spoken performance) through a number of
biases, including a tendency to inflate scores for a
number of reasons, such as believing moderate grades
do not reflect the quality of the instruction or students,
a tendency to be a harsh or tough marker (or
conversely to be a lenient or soft marker), a tendency
to be inconsistent, perhaps through inattention or
fatigue, and a tendency to judge the learner rather than
the performance itself. Brown & Hattie, 2012, p. 288289 Indeed, inspection of essay marking at university
level education shows that it is notoriously unreliable
(Brown, 2009b). Furthermore, research studies of
student self-assessment show that it quite unreliable
for novice and low-performing students (Brown &
Harris, 2012). Hence, the judgments teachers make
on-the-fly contain error and any decisions based on
such judgments may be dangerous for learning. High
quality decisions depend on data obtained through
multiple methods, multiple opportunities to perform,
and multiple judges (Brennan, 1996).
Even though well-designed tests provide
estimates of error, this good work can be undone by
inadequate reporting of information to teachers and
students. Too often tests only provide total or rank
order scores, which as useful as they are, do not give
teachers, let alone students, the information they need
to identify who needs to be taught what next (Hattie &
Brown, 2010). Considerable effort went into the
design of reports in the Assessment Tools for
Teaching and Learning (asTTle) system (Brown,
2005; Hattie & Brown, 2008; Hattie, Brown, &
Keegan, 2003) and it was found that teachers who
endorsed the improvement purpose of assessment had
better understanding of the assessment reports (Hattie,
Brown, Ward, Irving, & Keegan, 2006). Field studies
have found that teachers who use the asTTle test
system gained in their professional understanding of
content being assessed and their st Ց