The article is focused on the issue of understanding social cognitive mapping (SCM) and analyzing whether it is a black box method as claimed by Farmer and Xie, or not. The authors try to demonstrate empirically what SCM measures and according to their view, it measures peer interactions. They also state that Farmer and Xie’s suggestion that the probe measures developmental processes, perceptions and shared preferences is incorrect, as such wide range of different phenomena cannot be measured by such a simple method. Neal & Neal (2013) prove their argument by stating the fact that there are more precise and elaborate methods of measurement such difficult phenomena.
The empirical validity of the social cognitive mapping also has been put into question. Previous SCM procedures were validated on the basis of peer interactions that Farmer and Xie consider as a lack of empirical validation. Therefore, they claim it to be false. However, referring to Gest et al. (2003), usage of co-nomination matrix by Neil & Neil (2013) in their article serves as enough proof of empirical validity. Moreover, the authors of the article at hand also rely on the years of sociological research that proves the validity of the social cognitive mapping. In terms of empirical sensitivity, SCM has also been criticized by Farmer and Xie. However, Neil & Neil (2013) acknowledge that the traditional SCM indeed needs more empirical sensitivity in order to test whether teens synchronize their behaviors. In addition, the authors of the article also recognize that their examples are very hypothetical and designed in order to detect limitations of SCM. Neil & Neil (2013) believe that unless boundary conditions of SCM validity are clearly defined, there is no point in all attempts to get more accurate results.
Farmer and Xie also critique in terms of errors and interaction groups as they believe that representatives of the sample may not be reliable because they can refer to the same real social group. Neil & Neil (2013) fairly notice that no one can be sure about errors and omissions during SCM analysis. For this purpose, co-nomination matrix is developed in order to neglect the effect of the errors and omissions to a certain extent. The article also does not assume that individual nominations reflect accurate real social groups.
SCM is often referred as to a black box method that helps identifying different peer groups in the sample. Farmer and Xie seem to agree with this as they used this three times in their commentary. However, Neil & Neil (2013) do not completely agree with this fact because there is some ambiguity regarding this issue. The main argument is that nobody knows for sure how the SCM algorithm works. The loss of information should be also taken into account as it produces lots of problems regarding the use of data. Moreover, interpretation of co-nomination matrix is crucially important for the SCM analysis to be comprehensive. Farmer and Xie misinterpret the matrix and this indeed makes some details about SCM hidden in the black box. Therefore, peer interactions and groups on their basis are not completely understood.
The similarity matrix is also a subject of criticism. Neil & Neil (2013) insist on its accuracy, but thanks to Farmer and Xie they found another problem. The authors came to conclusion that the off-diagonal cells should be the element of the similarity matrix in order to reflect correct relationships of co-nominators. Farmer and Xie’s similarity matrix was constructed taking both off-diagonal and diagonal cells and this led to some incorrect outcomes. They defend their approach by explaining some psychological peculiarities of youth that constituted the sample.
Moreover, Neil & Neil (2013) in their original article developed an SCM program that is dedicated to help researchers in this field. It computes, construct matrixes, and normalize co-nomination in order to make the social development study easier and more informative.
References
Neal, Z. P. & Neal, J. W. (2013, August). Opening the Black Box of Social Cognitive Mapping. Social Development, 22(3), 604-608.
