The proposed framework represents a staticrnanalysis framework to classify the Android malware. From each Android .apkrnfile, three distinct features likely (a) opcodes (b) methods and (c)rnpermissions are extracted. Analysis of Variance (X-ANOVA) is used to rankrnfeatures that have high difference in variance in malware and benign trainingrnset. To achieve this conventional ANOVA was modified; and a novel techniquernreferred to us as X–ANOVA is proposed. Especially, X–ANOVA is utilized tornreduce the dimensions of large feature space in order to minimizernclassification error and processing overhead incurred during the learningrnphase. Accuracy of the proposed system is computed using three classifiersrn(J48, ADABoostM1, Random Forest) and the performance is compared with votedrnclassification approach. An overall accuracy of 88.30% with opcodes, 87.81%rnwith method and an accuracy of 90.47% is obtained considering permission asrnfeatures, using independent classifiers. However, using voted classificationrnapproach, an accuracy of 88.27% and 87.53% are obtained respectively forrnfeatures like opcodes and methods. Also, an improved accuracy of 90.63% wasrnascertained considering permissions. Initial results are promising whichrndemonstrate that the proposed approach can be used to assist mobilernantiviruses.