Policy and Complex Systems
networks emerging and students actively instructing other students ( Johnson , 2016 ).
Phase IV results . The results of Phase IV were developed from the data from Phase I , Phase II , and Phase III . The data-driven ABM measured and showed the changes in the types of interaction networks in classroom environments over half hour time periods and predicted average classroom achievement scores . The intervention School # 1 exhibited the highest level of change in positive classroom environment . This can be seen in Figure 3 by the yellow squares that represented positive CLASS behavioral markers that contribute to improving achievement ( CASTL , n . d .). Blue squares represent CLASS behavior markers that detract from a positive environment . The data for predictions were drawn from the random selection of individual student achievement scores from 2013 ( when the common core was initiated ) through 2015 . The ABM in Class # 1 predicted the highest improvement in advanced mathematics classes at 110 % or 11 points on a scale score for achievement ( see Figure 3 ). School # 2 ’ s prediction was a 3 % improvement . For School # 3 observational date and parameter ranges were substituted with similar profiles from previous research ( Johnson , 2015 ) and had predictions of 67 % improvement for regular mathematics classes and 83 % for advanced ( Johnson , 2016 ). The ABM adheres to the widely utilized overview design concepts and details ( ODD ) protocol by Railsback and Grimm ( 2012 ) and meets recommended ABM standards of validity and validation by Macal ( 2005 ) ( see Appendix C ).
Phase V results . The results of Phase V , based on the data from Phase I , Phase II , Phase III , and Phase IV , showed that the integrated methodology could account for the variety of systems such as instructional , emotional , behavioral , climate , structural , hierarchical , and even chaos in the network analysis . Further , the interaction patterns contributed to cocreating the classroom climate , which led to high Teacher Control , which led to effective teaching for Common Core second semester , which led to improvements that ranged from 11 % to 130 %. The ABM predicted scores for 2014-2015 from simulation runs ranged from 93.6 % to 100 % accuracy ( Johnson , 2016 ). Table 3 shows a comparison of the results .
Interpretations
This study presented multiphased
robust results explicating fundamental classroom interactions , classroom climate , and a predictive achievement score model to help plan for a more active participatory role in the creation of more positive futures for students .
Fully comprehending a policy reform , such as charter schools , requires a systems approach contextualized within a theoretical complexity context , which was accomplished in this research ( Lasky , 2004 ; Patton , 1990 ). The complexity context and methodologies allowed for capturing more and nuanced levels of system details . For example , results were presented on varied scales : individual students , groups of students , individual teachers , classrooms , and schools . These results are consistent with the claim by Kaput et al . ( 1999 ) that a complexity context facilitates understanding systems and changes in education in relation to a policy reform strategy ( Johnson , 2016 ).
The ABM simulations enabled visual representations of classroom interactions , networks , and climate with aggregate classroom and individual agent
35