On 2017 May 25, Lydia Maniatis commented:

"A simple alternative model is developed that is consistent with the results."

I.e. an ad hoc, data-fitted proposal is submitted that, as such, is experimentally uncorroborated and carries no theoretical weight. This is because experimental conditions must always be informed by theoretical assumptions (which they are supposed to test), so that the experiment can select which variables to hold constant, and which to allow to vary so as to test their hypothesized role, Post hoc explanations cannot distinguish between causal variables and confounds, and thus the experiments which they are evaluating in hindsight cannot serve as tests of a hypothesis so-derived.

I suppose the authors will test their proposal in the future, but at the the present stage it was not worth reporting. Testing it would, of course, require much more extensive and detailed conceptual development, precisely so that anyone wanting to test it could exercise the necessary experimental control over the presumed causal variables, and minimize researcher degrees of freedom, i.e. the degree of uncertainty, in interpreting results.

On 2017 May 25, Lydia Maniatis commented:

"A simple alternative model is developed that is consistent with the results."

I.e. an ad hoc, data-fitted proposal is submitted that, as such, is experimentally uncorroborated and carries no theoretical weight. This is because experimental conditions must always be informed by theoretical assumptions (which they are supposed to test), so that the experiment can select which variables to hold constant, and which to allow to vary so as to test their hypothesized role, Post hoc explanations cannot distinguish between causal variables and confounds, and thus the experiments which they are evaluating in hindsight cannot serve as tests of a hypothesis so-derived.

I suppose the authors will test their proposal in the future, but at the the present stage it was not worth reporting. Testing it would, of course, require much more extensive and detailed conceptual development, precisely so that anyone wanting to test it could exercise the necessary experimental control over the presumed causal variables, and minimize researcher degrees of freedom, i.e. the degree of uncertainty, in interpreting results.