On 2016 Jan 12, Elaine Beller commented:

This is a topic that has been debated on my Facebook feed recently. This group has performed a RCT, but reports within-group changes throughout, and not the between-group results. Some of the within-group results have P-values less than 0.05, and these are the ones that been reported in the abstract. This approach implies that there were no changes in the control group, and that both groups had similar baseline values (which is not so). I suggest the authors should give the results for the comparison between groups on final value, adjusted for baseline (ANCOVA).

Bland and Altman have written several excellent artcles explaining why the within-groups approach should not be used. In one which is open-access, they say: "This approach is biased and invalid, producing conclusions which are, potentially, highly misleading. The actual alpha level of this procedure can be as high as 0.50 for two groups...",

Comparisons against baseline within randomised groups are often used and can be highly misleading. Trials 2011, 12:264 doi:10.1186/1745-6215-12-264

On 2016 Jan 12, Elaine Beller commented:

This is a topic that has been debated on my Facebook feed recently. This group has performed a RCT, but reports within-group changes throughout, and not the between-group results. Some of the within-group results have P-values less than 0.05, and these are the ones that been reported in the abstract. This approach implies that there were no changes in the control group, and that both groups had similar baseline values (which is not so). I suggest the authors should give the results for the comparison between groups on final value, adjusted for baseline (ANCOVA).

Bland and Altman have written several excellent artcles explaining why the within-groups approach should not be used. In one which is open-access, they say: "This approach is biased and invalid, producing conclusions which are, potentially, highly misleading. The actual alpha level of this procedure can be as high as 0.50 for two groups...",

Comparisons against baseline within randomised groups are often used and can be highly misleading. Trials 2011, 12:264 doi:10.1186/1745-6215-12-264