METHODOLOGICAL CRITICISM OF A SCIENTIFIC ARTICLE
«Dessin de la main et habiletés numériques chez les enfants d’âge préscolaire
Bonneton-botté, N., Hilli, H., De La Haye, F. & Noêl, Y. (2015)»
The authors of this article dealt with the relation between manual representations and digital skills. The scientific domain was not able to choose between a functional and locational hypothesis of this relation. The authors suppose that there is a functional link between them and thus, that a training in the skills of manual representation should have a positive impact on the digital competence. The first objective observe and describe the children’s ability to draw a model-free hand during the fifth year. The authors’ second aim derives directly from the first. They seekto describe the relationships that characterize hand drawing performance and digital proof performance in the same child.
This quasi-experimental design has several strengths: regarding the choice of participants,to have a better representation, a fair division was divided between girls and boys composed each of 31 subjects.Some inclusion and exclusion criteria have also been added. The participants are all from periurban schools. The studyexcluded participants with proven disorders and developmental delays to control for these variables.The tests took place in the morning, which made it possible to encourage students’ concentration since they could not then be mobilized by past events during the day. Moreover, the tests took place in a neutral space in the school environment, which is an asset because the participants were in a favorable environment, compared perhaps to a laboratory that could have provoked nervousness, because it’s a known place.
Regarding weaknesses, few variables were checked to ensure internal validity. An addition of a motivation evaluation grid could have been added in order to be able to control for external variables that could influence the main variables.The sample chosen is small, consisting of only 62 participants, which decreases the representativeness of the sample. It will therefore be difficult to extrapolate the results collected to a larger population. A sample with a larger number of participants such as one hundred could have been more representative in order to extrapolate the results on the population of Grade 5 students.The instructions given during the drawing of the hand as well as during the numerical skills tests were issued by the same experimenter with the same words in a clear and precise way, this is not a problem. On the other hand, a better protocol would target a better measurement tool. Another more appropriate measurement tool could have been used to measure the representation of the children’s hand than a drawing of it. A specialized equipment could be used such as an ipad with an application that would allow participants to perform a hand representation in 3D to see if they have a good manual representation. This would allow control other factors such as the skill of art. With this application we could exclude that children less adept in drawing do not score well in manual representation for lack of skill.
The proposed method meets the objectives (ability to draw a hand, relationship between drawing performance and digital proofs) because of its tests, indeed they refer to what the authors want to observe.The first corresponds to the drawing of the hand, and to better evaluate this skill of the child, several elements were taken into account in this measure as the quality of the drawing, number of fingers and planning, so it measures the drawing well from the hand. The second test consists of two tests, one of term-to-term correspondence and the other of non-relevance of the order, these two age-appropriate digital proofs and the children’s skills are therefore useful to represent the acquisitions.(In addition, children who had a better strategy and planning for drawing were more successful in digital tasks, proving the relationship between these two skills).
The method is therefore appropriate for the scientificdesign, that is quasi-experimental, the method was made with a non-random selection of participants. The experiment is done in a school environment and not a laboratory. Experimenters were able to control the timing of treatment exposure in the morning.Quasi-experimental design has also allowed for multiple assessments as in the case of numerical skills tests that are part of the method.
The choice of thescientificdesignis a weakness in itself, because the independent variable can’tbe manipulated. Validity is therefore threatened since it does not allow us to reject other explanations of the relationship between the independent and dependent variable.An experimental design with a control group and an experimental group could have been more appropriate. This would have made it possible to see not only if a relationship exists between the variables but also to see how the variables influence each other and to discard other explanations of the relationship.
Regarding ethical and deontological approaches, there is no mention of how the children were recruited for the study, did the parents agree? Was tinformedor forced voluntary participation? Nor is there any mention of whether the children or parents were well informed about the aims of the study and why they participated. Were there X risks for the subjects of the study, such as self-esteem and feeling of effectiveness? We can only assume that these ethical procedures were applied by researchers because they had to go before an ethics committee to get their research accepted. However,we must give credit to the morality of the authors, indeed despite some more or less conclusive results, they have not tried to falsify them to make excellent results that are consistent with all their assumptions. On the contrary, they have clearly shown at the end of the article, in the discussion section, what these inconclusive results nevertheless bring to science.
Keywords: Digital competence; manual representation; method; quasi-experimental design; deontology