colleagues and I have drawn distinctions between two forms of subject-
matter knowledge—domain and topic knowledge (e.g., Alexander & Kuli-
kowich, 1994). Domain knowledge represents the breadth of one’s knowl-
edge, whereas topic knowledge deals with understanding of a particular topic
or concept within that field.
Thus, in Alexander, Kulikowich, and Schulze (1994), we tested undergrad-
uates’ knowledge of astrophysics by means of a multiple-choice test covering
a range of astrophysics content, and also focused on students’ depth of un-
derstanding about certain topics from that field, such as black holes, quarks,
or Stephen Hawking. We considered it important to examine both forms,
given that those relatively new to an academic domain may not know a great
deal about the domain of astrophysics but they may know something about
select topics in that domain. In fact, in Alexander et al. (1989), we found that
elementary students could demonstrate knowledge of certain biological con-
cepts (e.g., heart) without understanding how those concepts related to bio-
logical systems (e.g., circulatory) or to the domain itself.
Strategies. The second dimension examined in the MDL is strategic
processing, which relates to both general cognitive procedures used in task
performance (e.g., summarization) and metacognitive strategies (e.g., self-
testing or self-evaluation) pertaining to the monitoring or regulation of one’s
learning (e.g., Garner & Alexander, 1989; Weinstein & Mayer, 1986). Do-
main-specific strategies, which are similarly critical to expert performance,
are incorporated in the MDL as part of domain knowledge. Initially, I hy-
pothesized a single curvilinear trajectory for strategic processing across the
stages of domain learning, disregarding the possibility of quality shifts in
strategy use (Alexander, 1997). My colleagues and I later determined that
such an approach masked important developmental differences and did not
allow the shifting relations between strategy use and knowledge to manifest
(Alexander & Murphy, 1998).
Thus, in more recent investigations, my fellow researchers and I have
sorted strategies into two classes—surface level and deep processing (Alex-
ander, Sperl, Buehl, Fives, & Chiu, 2002; Murphy & Alexander, 2002; Van-
Sledright & Alexander, 2001). This classification has resulted in a clearer
distinction between novices and more competent or more proficient learn-
ers. Surface-level reading strategies, such as rereading or omitting unfamil-
iar words, facilitate the initial comprehension of the domain text. In effect,
these strategies give readers access to the message. By comparison, deep-
level processing strategies, such as relating text to prior knowledge or ques-
tioning the author, involve the personalization or transformation of the
message.
We have also included what we call interactive measures in recent studies
(Alexander et al., 2002; Murphy & Alexander, 2002). These measures permit
- MODEL OF DOMAIN LEARNING 285