Web.cs.dal.ca

Enter raw data from SPSS or SAS into RED cells then click a YELLOW box to calculate:GROUP #1:MEAN15(S) STANDARD DEVIATION 2n50GROUP #2:MEAN12(S) STANDARD DEVIATION2n50**ALWAYS CLICK HERE TO CALCULATE FINAL RESULTS**Effect Size:=IF(ISNUMBER(#REF!),IF(ISNUMBER(B7),IF(ISNUMBER(B12),IF(ISNUMBER(B13), B15-CONFIDENCE(0.05,#REF!,(B9+B14)),""),""),""),"")=IF(ISNUMBER(#REF!),IF(ISNUMBER(B7),IF(ISNUMBER(B12),IF(ISNUMBER(B13), B15+CONFIDENCE(0.05,#REF!,(B9+B14)),""),""),""),"")95% Confidence Intervals:LowerUpperd Effect Size (Maximum Likelihood Estimator)*=IF(ISNUMBER(B22),(B5-B10)/B22,"")=IF(ISNUMBER(B18),IF(ISNUMBER(B5),IF(ISNUMBER(B10),IF(ISNUMBER(B11), B18-CONFIDENCE(0.05,B22,(B7+B12)),""),""),""),"")=IF(ISNUMBER(B18),IF(ISNUMBER(B5),IF(ISNUMBER(B11),IF(ISNUMBER(B10), B18+CONFIDENCE(0.05,B22,(B7+B12)),""),""),""),"")Standardized Mean Difference*=IF(ISNUMBER(B23),(B5-B10)/B23,"")Hedge's g (unbiased) =IF(ISNUMBER(B23),((1 - (3 / (4 * (B7+B12-2) - 1)))*B19),"")=IF(ISNUMBER(B20), B20-CONFIDENCE(0.05,B23,(B7+B12)),"")=IF(ISNUMBER(B20), B20+CONFIDENCE(0.05,B23,(B7+B12)),"")g converted to r Effect Size=IF(ISNUMBER(B20), (B20)/(SQRT(B20^2+4*((B7+B12-2)/(B7+B12)))),"")=IF(ISNUMBER(B21),IF(ISNUMBER(B5),IF(ISNUMBER(B10),IF(ISNUMBER(B11), B21-CONFIDENCE(0.05,B22,(B7+B12)),""),""),""),"")=IF(ISNUMBER(B21),IF(ISNUMBER(B5),IF(ISNUMBER(B11),IF(ISNUMBER(B10), B21+CONFIDENCE(0.05,B22,(B7+B12)),""),""),""),"")σ pooled (pooled standard deviation)=IF(ISNUMBER(B23),B23*(SQRT((B7+B12-2)/(B7+B12))),"")=IF(ISNUMBER(B17),IF(ISNUMBER(B10),IF(ISNUMBER(#REF!),IF(ISNUMBER(#REF!), B22-CONFIDENCE(0.05,B16,(B12+#REF!)),""),""),""),"")=IF(ISNUMBER(B17),IF(ISNUMBER(B10),IF(ISNUMBER(#REF!),IF(ISNUMBER(#REF!), B22+CONFIDENCE(0.05,B16,(B12+#REF!)),""),""),""),"")S pooled (pooled standard deviation)=IF(ISNUMBER(B12),(SQRT(((B7-1)*(B6^2)+(B12-1)*(B11^2))/(B7+B12-2))),"")(The Effect Sizes calculated uses standard deviation and are appropriately weighted when the sample sizes entered are not equal.)*The Standardized Mean Difference has been labeled as Cohen's d by some authors or as Hedge's g by others. Hedge's g (unbiased) is the more conservative calculation of Effect Size.EFFECT SIZE CONVERSIONS:To convert d Effect Size into Hedge's g Effect Size, enter n sizes (if available)* and d value into the RED cells, then click the YELLOW box to calculate:To convert Hedge's g Effect Size into Cohen's d Effect Size, enter n sizes (if available)* and d value into the RED cells, then click the YELLOW box to calculate:Group #1 nEnter #Group #1 nEnter #Group #2 nEnter #Group #2 nEnter #d Effect SizeEnter #g Effect SizeEnter #**CLICK HERE TO CALCULATE FINAL RESULTS****CLICK HERE TO CALCULATE FINAL RESULTS**d Effect Size converted to g=IF(ISNUMBER(B37), B37*(SQRT((B35+B36-2)/(B35+B36))),"")g Effect Size converted to d=IF(ISNUMBER(F37), F37*(SQRT((F35+F36)/(F35+F36-2))),"")To convert r Effect Size into d Effect Size, enter n sizes (if available)* and d value into the RED cells, then click the YELLOW box to calculate:*If n sizes are not provided, the r->d conversion formula provides an estimation assuming EQUAL sample sizes.To convert d Effect Size into r Effect Size, enter n sizes (if available)* and d value into the RED cells, then click the YELLOW box to calculate:*If n sizes are not provided, the d->r conversion formula provides an estimation assuming EQUAL sample sizes.Group #1 nEnter #Group #1 nEnter #Group #2 nEnter #Group #2 nEnter #r Effect SizeEnter #d Effect SizeEnter #(enter r as a number between 0 and 1)**CLICK HERE TO CALCULATE FINAL RESULTS****CLICK HERE TO CALCULATE FINAL RESULTS**d Effect Size converted to r=IF(ISNUMBER(F47),IF(ISNUMBER(F46),(F47)/(SQRT(F47^2+(((F45+F46)^2))/(F45*F46))),F47/((F47^2+4)^(1/2))),"")r Effect Size converted to d=IF(ISNUMBER(B47),IF(ISNUMBER(B46),(B47*(B45+B46))/(SQRT(B45*B46*(1-B47^2))),2*B47/SQRT(1-B47^2)),"")=IF(ISNUMBER(B5),IF(ISNUMBER(B6),IF(ISNUMBER(C5),IF(ISNUMBER(C6),IF(ISNUMBER(#REF!),(C6)/(C5+C6),""),""),""),""),"")To convert an independent samples t-score into r and d Effect Sizes, enter raw data in RED cells, then click the YELLOW box to calculate:To convert an F-score (that is a comparison of two independent groups) into r and d Effect Sizes, enter raw data in RED cells, then click the YELLOW box to calculate:Group #1 nEnter #**NOTE:The F-value has to be from a comparison of just two groups. Conversions to r or d should not be attempted from an omnibus F test (i.e., with numerator df > 1).Group #2 nEnter #t-scoreEnter #**CLICK HERE TO CALCULATE FINAL RESULTS**t-score converted to r Effect Size=IF(ISNUMBER(F58),IF(ISNUMBER(F57), SQRT((F58^2)/((F58^2)+(F56+F57-2))),""),"")Group #1 nEnter #t-score converted to d Effect Size=IF(ISNUMBER(F58),IF(ISNUMBER(F57),((F58*(F56+F57))/(SQRT(F56+F57-2)*SQRT(F56*F57))),""),"")Group #2 nEnter #F-valueEnter #**CLICK HERE TO CALCULATE FINAL RESULTS**degrees of freedom within=IF(ISNUMBER(B61),IF(ISNUMBER(B62),B61+B62-2,""),"")To convert a dependent samples t-score (i.e. pre/post-test design) into r and d Effect Sizes, enter raw data in RED cells, then click the YELLOW box to calculate:F-value converted to r Effect Size=IF(ISNUMBER(B63),IF(ISNUMBER(B65),SQRT((B63)/(B63+B65)),""),"")F-value converted to d Effect Size=IF(ISNUMBER(B63),IF(ISNUMBER(B61),IF(ISNUMBER(B62),IF(ISNUMBER(B65),(SQRT(B63)*(B61+B62))/(SQRT(B65)*SQRT(B61*B62)),""),""),""),"")To convert an F-score (that is a Pre/post comparison) into r and d Effect Sizes, enter raw data in RED cells, then click the YELLOW box to calculate:Total NEnter #t-scoreEnter #**CLICK HERE TO CALCULATE FINAL RESULTS**t-score converted to r Effect Size=IF(ISNUMBER(F71),IF(ISNUMBER(F70),SQRT((F71^2)/((F71^2)+(F70-1))),""),"")**NOTE:The F-value has to be from a comparison of just two groups. Conversions to r or d should not be attempted from an omnibus F test (i.e., with numerator df > 1).t-score converted to d Effect Size=IF(ISNUMBER(F71),IF(ISNUMBER(F70),((2*F71)/(SQRT(F70-1))),""),"")To convert a 2x2 Chi-Square into r and d Effect Sizes, enter raw data in RED cells, then click the YELLOW box to calculate:Total NEnter #F-valueEnter #**CLICK HERE TO CALCULATE FINAL RESULTS****NOTE:The chi-square has to be from a comparison of just two groups with two dimensions (2 x 2). Conversions to r or d should not be attempted from an omnibus chi-square test (i.e., with df > 1).degrees of freedom within=IF(ISNUMBER(B78), B78-1,"")F-value converted to r Effect Size=IF(ISNUMBER(B79),IF(ISNUMBER(B81),SQRT((B79)/(B79+B81)),""),"")F-value converted to d Effect Size=IF(ISNUMBER(B78),IF(ISNUMBER(B79),((2*SQRT(B79))/(SQRT(B81))),""),"")Total # (N)Enter #Chi-square valueEnter #*Program created by Jared DeFife, Ph.D., Emory University, 2009**CLICK HERE TO CALCULATE FINAL RESULTS**Chi-square converted to r Effect Size=IF(ISNUMBER(F85),IF(ISNUMBER(F84),SQRT((F85)/(F84)),""),"")Chi-square converted to d Effect Size=IF(ISNUMBER(F85),IF(ISNUMBER(F84),2*SQRT((F85)/(F84-F85)),""),"") ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download