TABLE 10.1(Continued)Alexander,Kulikowich, &
Schulze (1994)Alexander, Jetton,& Kulikowich(1995)Alexander, Murphy,Woods, Duhon, &Parker (1997)Alexander &
Murphy (1998)Murphy &
Alexander (2002)Alexander, Sperl,Buehl, Fives, &Chiu (2002)AnalysesMANOVA; Re-gressionMANOVA; ClusteranalysisMANOVA; SEMCluster analysisMANOVA;PathanalysesMANOVA; ClusteranalysisFindingsDomain knowledgepredicted stu-dents’ recall andinterest; Rela-tions betweendomain knowl-edge, interest,and recall grewstronger acrossperformancegroupsExp. 1: Three dis-tinct clustersvarying byknowledge, inter-est, and recall
Exp. 2: Four clus-ters ranging fromhigh knowledge,interest, and re-call to low on allvariablesSignificant in-creases in knowl-edge and interestover time anddecreased use oftext-based (sur-face-level) strate-gies; expected re-lations betweenand amongmodel factorsThree distinct clus-ters formed atpretest and fourclusters wereidentified atposttest. Charac-teristics of clus-ters confirmedstrongly tomodel predic-tions.Path analysesshowed thatposttest subject-matter knowl-edge was directlyand indirectlypredicted by pre-test knowledge,surface- anddeep-level strate-gies, interactiveknowledge, andpretest interestFour clustersemerged with thecharacteristics ofacclimation, mid-competence,high-competence,and proficiency.288
