These are used as a first step to quickly compare results from different groups positive deviants non positive deviants average performers or negative deviants This provides a basic overview of the data e g mean standard deviation min and max Quantitative data analysis Descriptive statistics They are used to identify significant differences between the means of the analyzed groups for each variable using statistical tests like student t test Mann Whitney and Fisher exact test when comparing two groups and tests like ANOVA Kruskal Wallis and Chi square when comparing three or more groups e g PDs average performers and NDs Inferential statistical tests This analysis is used to examine the relationship between the captured performance measure dependent variable and the independent variables Regression analysis enables you to estimate the effect of each independent variable on the dependent variable while holding all the other independent variables constant In this stage the independent variables will include both the structural variables used in the positive deviant identification step and the sociodemographic and behavioral variables captured in the data collection step For each homogeneous group you will have a separate model to identify significant predictors of performance that are relevant to the respective group Regression analysis You might consider regressing the performance measure using different subsets of the independent variables For example modeling the performance measure as a function of structural variables that are beyond the level of individuals modeling separately behavioral predictors of performance then creating a full model containing both subsets This way you might be able to understand the effect of each subset of variables separately which could guide the design of the interventions If structural variables have more influence then you might consider a systematic policy focused intervention If the behavioral variables stand out then you should consider a community focused intervention If both have significant effects then you should consider an intervention combining both Cluster then predict Check this guide for further guidance on which test to use https www scielo br j mioc a qMBfF39f568Wm8gyB3YJmbH abstract lang en 93 94

Vorschau DPPD Handbook Seite 48
Hinweis: Dies ist eine maschinenlesbare No-Flash Ansicht.
Klicken Sie hier um zur Online-Version zu gelangen.