Electricity Use Behaviour in a High-Income Neighbourhood in Johannesburg, South Africa

Round 1

Reviewer 1 Report

The paper explores electricity use behavior among high-income households in Johannesburg, South Africa in 2017. The data was collected using structured questionnaires and several different non-parametric tests, multivariate regression and factor analysis were conducted to test the validity of the hypotheses.  Bearing in mind that little is known about electricity use behavior in South Africa and that its understanding is important for formulating policy interventions, the subject is actual and interesting. However, my major concerns related to this paper are of a methodological nature.

• The authors explained that they follow the previous literature in creating latent, i.e., personal value variables. However, information is needed with regards to manifest variables that were used for their creation and measuring as well as to this process itself.
• A data screening procedure should be conducted before any statistical analyses. Was it conducted?
• In order to obtain credible results from multivariate regression and factor analysis, certain assumptions must be satisfied and diagnostic tests conducted. The authors did not report that such assumptions are met and the diagnostic tests performed. For example, the initial assumption to employ these methods is to have a minimum sample size. Could 91 observations ensure credible results? Or, for example, is the issue of multicollinearity addressed in this paper?
• In addition, more information should be provided for factor analysis. For instance, the method used for the factor extraction and rotation should be noted.

Minor remarks

• 26 electricity use actions are announced to be used to assess electricity use behavior in Subsection 2.2. The authors can direct the readers to Table 2 (the first column) where they are listed. The same case holds for personal value variables.
• The sample size is one of the limitations of the paper and this should be added to the last section.
• The brief contribution of the paper and its organization should be given in the introduction.
• The number of observations should be added to Tables 1 and 2.
• To enhance readability of the paper, more paragraphs should be added in Section 4.1.
• Since the readers of the Sustainability journal are international, it would be better if the average national household income (p. 6) is expressed in some international unit.
• The acronym BRICS should be explained at the first mention (p. 2).
• In many countries the share of households in final energy consumption accounts for more than 20% (p. 1)
• The first sentence in the introduction section should be reformulated. It implicates that electricity use generates mostly negative impacts for human well-being.

Author Response

Thank you for the useful comments. We have attempted to address all the comments. Please, see attached response document.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper provides some interesting information about the electricity use behaviour of the affluent society of Johannesburg. The data collection and analysis is fairly straightforward and groupings of various factors is also useful in interpreting the findings. I believe the conclusions will be use useful to the policymakers.

 

As with any other structured survey, the collected data forms the crucial element of this research. I would like to see more information on the data collection logistics. In particular:

• Specific information about how many data were collected face to face and how many were supplied by the household occupants should be noted.
• It would be useful and informative for any future researchers to know the logistics of how a good number of interviews were conducted over a two-month period by one researcher.

Author Response

Thank you for the useful comments. We have addressed these in the attached response document.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The paper explores electricity use behavior among high-income households in Johannesburg, South Africa in 2017. The data was collected using structured questionnaires and several different non-parametric tests, multivariate regression and factor analysis were conducted to test the validity of the hypotheses.  Bearing in mind that little is known about electricity usage behavior in South Africa and that its understanding is important for formulating policy interventions, the subject is actual and interesting.

The authors addressed most of my doubts related to their paper. However, my major doubt related to the sample size should be more carefully considered.

 

Certainly, it is true that a small sample can be very beneficial in providing useful information about the subject matter, but the methods applied to analyze such sample should be fitted to its size. Hence, I do not question the content and the purpose of this paper, but the choice of the methods in the analysis of the collected data. A multiple regression cannot be applied when a sample is too small. The result of the F-test in this paper indicates an insignificant model and results. Hence, its results should not be the basis for conclusion drawing. Besides, the results of other diagnostic test for multiple regression are not provided in the paper (e.g.,  autocorrelation). Moreover, it is not enough to write that VIFs are not an issue in the paper; their scores should be given in the appendix. The same situation regarding the sample size issue holds with factor analysis. Does the sample size ensure reliable results in this paper? This is the question that needs to be answered in this paper and the answer very carefully and appropriately commented and supported with the references.

Author Response

Thank you for raising pertinent questions. We have attempted to address these below and hope the concerns are now sufficiently addressed. However, we are happy to address these further if needed, but kindly ask the reviewer to provide specific suggestions on appropriate statistical techniques we should consider.

 

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

I have only one minor remark:the title “Understanding electricity use habits and behavior” should become Section 2, “Materials and methods” Section 3, etc.

Author Response

Section numbereing has been revised as per the reviewer's suggestion.