Safety Climate in the Indonesian Construction Industry: Strengths, Weaknesses, and Influential Demographic Characteristics

Round 1

Reviewer 1 Report

Comment #1: This study used a comparative analysis to analyze the differences in safety climate level based on respondents’ demographic. This study used a sample of 1757 construction workers.  The title and abstract describe the contents of the paper correctly. Although this research provides valuable results for the specific country context, the reviewer has the following comments:

Comment #2: There are a recent studies of construction safety climate in Indonesian construction industry (see for example “Lestari et. al. (2020). A safety climate framework for improving health and safety in the Indonesian construction industry”), What can you learn from this study, and how about the difference between your study and previous work specially in addressing your research questions (A, B, and D):

 

1. “What is the level of safety climate in the Indonesian construction industry?”
2. “What are safety climate dimensions that require urgent improvements? c)”
3. “What is the influence of demographic profiles on safety climate?”
4. “ How can safety climate in the Indonesian construction industry be improved?”

 

 

Comment # 3: Although the authors well addressed the survey analysis including the validity and reliability analysis, they should develop a section for methodology and not include the methods information in the result section.

 

 

Comment # 4: the majority of respondent were from Staff (about 857), are we talking about management staff here or workers?

 

 

Comment # 5: Authors stated that (in line 319-321) “Those with higher education levels perceived that the safety climate score is lower than those with lower education levels. In contrast, previous research found that higher educated workers express more positive perceptions regarding safety than their lower educated counterparts” please provide citation/s here to support this argument?

 

 

 

Comment # 7: The reviewer noticed that is super difficult to see a link between work location specially comparing working in office and project and the two environment is different and risks associated with bot is different as well?  So how authors can compare between staff working in office and workers in the cite in section 5.6 (line 354) ? I hope authors can reconsider adding this point in this research. 

Author Response

Many thanks for your review.

Please see the attached response letter.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript is quite interesting to read, an always relevant issue and presenting a very large sample.  There, as normal, some details that could be improved, namely:

• Clarification of the scope, since the authors only cover public contractor and, despite the large number of replies, it only captures 6 companies. It would be interesting to the reader and important to the research robustness to know how much this represents the construction sector in Indonesia (how many companies there are, how many are private and how many are public, how many workers there are in the construction sector, value of the projects construction sector vs the value of these companies, etc). The sample is large, but it is impossible to assess if it is representative or not and it may reflect the perspective of small fraction of the construction industry – I do not see the problem with this, but it must be clear to the reader
• There are some typos, that would require correction (e.g, line 228 dan instead of and)

 

However, my biggest concerns with the manuscript are:

• How were the Reliability and Validity assessed? To the best of my knowledge reliability requires repeating the questions in different moments and measure the degree of agreement and validity requires assessing the degree of agreement between two methods of measuring the same phenomenon and I can’t disclose from the manuscript if this was done and how was it done. Perhaps the authors are using a different definition of reliability and validity, but this would require a clear definition of them in the paper.
• The authors treat the replies from the questionnaires as continuous when they aren’t – they are nominal (in fact, they are conversion of pure categorical variables into numbers, not even numeric categories such as age intervals). This, from a statistical point of view, is a major error that permeates over all the manuscript:
  • categorical variables are best categorized by cross-tabs and not mean and standard deviation;
  • it makes no sense to test the normality of a categorical variable because it is not applicable (normal distribution implies a continuous variable);
  • the recommended test for comparing groups are not the parametric or non-parametric means comparison tests because mean is not even a good measure of central tendency for categorical or nominal data.

 

I hope that the authors understand that, in its current form, the manuscript is not acceptable for publication because it is supported in a statistical analysis that is based in wrong assumption, so all calculations are not valid. The authors do not refer the software used for the statistical analysis, but consulting the online free guide of one of the most used software packages (SPSS) will clarify why are all the analysis wrong.

 

I am aware that the approach used by the authors has been used several times in the past in other research efforts. However, this doesn’t make it correct and it is only a reflexion that the reviewers might be experts in health and safety in construction but had limited knowledge in statistics. In the end, the manuscript is 80% about statistics and all the results and conclusions are made based on a wrong statistical analysis.

Author Response

Many thanks for your review.

Please see the attached response letter.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for addressing all comments