Which Skills Are the Most Absent among University Graduates in the Labour Market? Evidence from Slovakia

Round 1

Reviewer 1 Report

It has been a pleasure to review the manuscript “Which skills are the most absent among university graduates in the labour market? Evidence from Slovakia” (socsci-1907572).

This paper deals with skills mismatch in Slovakian labour market among university graduates. The work evaluates the satisfaction of both recent graduates and employers with the skills acquired during the studies. In their empirical analysis, the authors exploit an ad hoc survey comprising more than 13,000 graduates using and logit model. With the exception of graduates from technical universities, the results suggest that both graduates and employers are not satisfied with the result of the higher education studies. This particularly apply to foreign languages.

Overall, the paper is well written and, even though its degree of originality is moderate, it contributes to widen our knowledge of skills mismatch in Slovakia. I have only minor comments on the paper, which mainly refers to better contextualize several aspects of the research with appropriate references.

(1) There are some typos that could be addressed. E.g., line 149 “The The data,”.

(2) In the introduction, in line 22, at the end of the sentence “There are many papers published in other countries discussing the gap between 21 the skills that graduates acquire and those that employers require.”, it could be a good idea to include some general references such as McGuiness et al. (2018) and Muñoz de Bustillo et al. (2018).

(3) It could be interesting to include in the introduction a paragraph on the implications of educational mismatch on individuals’ well-being such as Badillo-Amador and Vila (2013) and Varona Cervantes and Cooper (2022). The authors could add that the effects of mismatch are likely to be more harmful for people with disabilities (Jones & Sloane, 2010), which usually experience a higher level of hardship (Antón et al., 2016). At the same time, I think that it could also be interesting to state that the consequences of skills mismatch can be even more relevant in the wake of the recent automation revolution (Fernández-Macías et al., 2020; Shmatko & Volkova, 2020).

References.

Antón, J.-I., Braña, F. J., & Muñoz de Bustillo, R. (2016). An analysis of the cost of disability across Europe using the standard of living approach. SERIEs: Journal of the Spanish Economic Association, 7(3), 281-306. https://doi.org/10.1007/s13209-016-0146-5

Badillo-Amador, L., & Vila, L. E. (2013). Education and skill mismatches: wage and job satisfaction consequences. International Journal of Manpower, 34(5), 416-428. https://doi.org/10.1108/IJM-05-2013-0116

Fernández-Macías, E., Klenert, D., & Antón, J.-I. (2021). Not so disruptive yet? Characteristics, distribution and determinants of robots in Europe. Structural Change and Economic Dynamics, 58, 76-89. https://doi.org/10.1016/j.strueco.2021.03.010

Jones, M. K., & Sloane, P. J. (2010). Disability and skill mismatch. Economic Record 86(10), 101-114. https://doi.org/10.1111/j.1475-4932.2010.00659.x

McGuinness, S., Pouliakas, K.; & Redmond, P. (2018). Overeducation in the Labour Market. Journal of Economic Surveys, 32(4), 985-1015-418. https://doi.org/10.1111/joes.12254

Muñoz de Bustillo, R., Sarkar, S., Sebastián, R., & Antón, J.-I. (2018). Skills mismatch: concept, measurement and policy approaches. International Journal of Manpower, 39(8), 977-995. https://doi.org/10.1108/IJM-10-2018-0327

Shmatko, N., & Volkova, G. (2020). Bridging the Skill Gap in Robotics: Global and National Environment. SAGE Open, 10(3), 1-13. https://doi.org/10.1177/2158244020958736

Santoso, G.. 2016. Technology As a Driver of Skills Obsolescence and Skills Mismatch: Implications for the Labour Market, Society and the Economy. ANU Undergraduate Research Journal, 7, 49-62. http://dx.doi.org/10.22459/AURJ.07.2015.06

Varona Cervantes, C., & Cooper, R. (2022). Labor market implications of education mismatch. European Economic Review, 148, 104179. https://doi.org/10.1016/j.euroecorev.2022.104179

Author Response

Author`s responses to Reviewer`s Comments

The authors are very grateful for the reviewer`s remarks concerning the submitted article. We appreciate the comments included in the review and find them highly valuable as they indicate the path for further research within the framework of our scientific activity. Please find below our responses:

Point 1: There are some typos that could be addressed. E.g., line 149 “The The data,”.

Response 1: The authors have made changes as suggested by the Reviewer.

Point 2: In the introduction, in line 22, at the end of the sentence “There are many papers published in other countries discussing the gap between 21 the skills that graduates acquire and those that employers require.”, it could be a good idea to include some general references such as McGuiness et al. (2018) and Muñoz de Bustillo et al. (2018).

Response 2: The authors would like to supplement the suggested references, but the articles refer to a slightly different topic - overeducation, which we do not deal with in our research. Nevertheless, thank you very much for the valuable articles, the quotations from which we use in the next article we are preparing.

Point 3: It could be interesting to include in the introduction a paragraph on the implications of educational mismatch on individuals’ well-being such as Badillo-Amador and Vila (2013) and Varona Cervantes and Cooper (2022). The authors could add that the effects of mismatch are likely to be more harmful for people with disabilities (Jones & Sloane, 2010), which usually experience a higher level of hardship (Antón et al., 2016). At the same time, I think that it could also be interesting to state that the consequences of skills mismatch can be even more relevant in the wake of the recent automation revolution (Fernández-Macías et al., 2020; Shmatko & Volkova, 2020).

Response 3: The authors have supplemented some of the recommended references which related to the reviewed topic, such as:  

• Badillo-Amador, L., & Vila, L. E. (2013). Education and skill mismatches: wage and job satisfaction consequences. International Journal of Manpower, 34(5), 416-428. https://doi.org/10.1108/IJM-05-2013-0116
• Shmatko, N., & Volkova, G. (2020). Bridging the Skill Gap in Robotics: Global and National Environment. SAGE Open, 10(3), 1-13. https://doi.org/10.1177/2158244020958736
• Varona Cervantes, C., & Cooper, R. (2022). Labor market implications of education mismatch. European Economic Review, 148, 104179. https://doi.org/10.1016/j.euroecorev.2022.104179

On the other hand the authors would like to supplement the suggested references, but the topics are slightly different. However, as we announced above, we are thankful to Reviewer for recommendation of these valuable articles which we use for for citation in our next upcoming article.

We would like to thank you for all suggestions regarding the improvement of the article quality. At the same time, we hope that it will be evaluated positively and accepted for publication in the journal.

Reviewer 2 Report

In the article titled "Which skills are the most absent among university graduates in the labor market? Evidence from Slovakia", it has been examined whether the employed people are satisfied with the programs they have studied and whether the most important factor affecting the satisfaction is employability. In addition, it was examined whether there was a difference between the skills acquired by graduates and the skills required by employers. The subject of the study is interesting. A literature review is sufficient. However, some of my criticisms and suggestions regarding the article are listed below.

 

1.       There are the Ho and H1 hypotheses in statistics. These significance values, which are valid in hypothesis tests, are determined as 1%, 5%, and 10%. Rarely, only one of them is used in the hypothesis tests performed at 10% in social sciences. Accordingly, Ho is accepted or rejected. It is scientifically incorrect to make qualifications such as important more important than in cases where the H1 hypothesis is statistically accepted, since there is no H2 hypothesis. One significant level is used in a scientific study. So my major comment on this article is related to this topic. According to basic statistical logic, the importance levels should be re-determined, and the interpretations should be made again based on these values. 

2.     The goodness of fit statistics are included in the article and are well explained. However, how were the variables determined during the model building phase? I couldn't find any explanation for this issue. It is important to look at at least VIF values from the assumptions about the logit model in determining the variables. In addition, post estimation is required for the logistic model in terms of classification and sensitivity.

3.       On page 9, comments are made with percentage values. However, logit model marginal effects are not mentioned anywhere or in the table of the article..

4.       The information between lines 213-219 can be given in smaller fonts under the table. In the section where the hypotheses are explained, 26 skills scaled from 1 to 10 are mentioned. It will be important to include the survey questions at the end of the article in order to see what they are as a whole and to understand what exactly is explained in the hypotheses. 

5.     The study consists of two parts in terms of analysis. However, only the logit model was mentioned in the material method section and the Wilcoxon signed-rank test was not included. After the statistical analysis comparing soft skills and the other, a discussion was made, but after the logit model, the current findings were not discussed with the previous studies. These sections should be made more regular in accordance with the flow of the article. However, another major criticism of this analysis, and therefore of the findings, is (as I mentioned earlier) that only one significance level is used in a scientific study. Therefore, either 5% or 10% significance level should be determined in Table 3.

6.     Instead of writing the p-value calculated in Gretl in the tables, it is sufficient to write only the p-value.

7.     The abstract section can be developed with a few sentences. The sentences should be more understandable and the article should be revised in terms of language

 

 

Author Response

Author`s responses to Reviewer`s Comments

The authors are very grateful for the reviewer`s remarks concerning the submitted article. We appreciate the comments included in the review and find them highly valuable as they indicate the path for further research within the framework of our scientific activity. Please find below our responses:

Point 1: There are the Ho and H1 hypotheses in statistics. These significance values, which are valid in hypothesis tests, are determined as 1%, 5%, and 10%. Rarely, only one of them is used in the hypothesis tests performed at 10% in social sciences. Accordingly, Ho is accepted or rejected. It is scientifically incorrect to make qualifications such as important more important than in cases where the H1 hypothesis is statistically accepted, since there is no H2 hypothesis. One significant level is used in a scientific study. So my major comment on this article is related to this topic. According to basic statistical logic, the importance levels should be re-determined, and the interpretations should be made again based on these values. 

Response 1: The authors have made changes as suggested by the Reviewer.

Point 2: The goodness of fit statistics are included in the article and are well explained. However, how were the variables determined during the model building phase? I couldn't find any explanation for this issue. It is important to look at least VIF values from the assumptions about the logit model in determining the variables. In addition, post estimation is required for the logistic model in terms of classification and sensitivity.

Response 2: On the basis of project cooperation, we participated in data collection within the national survey of university graduates. The variables of the model were built up in the frame of research project proposal by the international panel of experts. The project proposal including the model variables was approved by the opponents of the scientific grant agency.

We provided the calculation of VIF in Gretl; however the collinearity of variables was not confirmed. In Gretl the collinearity is given when VIF is higher than 10; our variables did not exceed the value of 4.7

We added sensitivity and specificity of logistic model in the notes under the table 1.

Point 3: On page 9, comments are made with percentage values. However, logit model marginal effects are not mentioned anywhere or in the table of the article.

Response 3: The authors have made changes and rewritten the interpretation as suggested by the Reviewer.

 Point 4: The information between lines 213-219 can be given in smaller fonts under the table. In the section where the hypotheses are explained, 26 skills scaled from 1 to 10 are mentioned. It will be important to include the survey questions at the end of the article in order to see what they are as a whole and to understand what exactly is explained in the hypotheses. 

Response 4: The authors have made changes as suggested by the Reviewer what about the fonts.

In relation to the survey, the scope of the questionnaire is very extend and is being used for another research documents and articles in the frame of research project for that reason, it would unduly lengthen the article.

 Point 5: The study consists of two parts in terms of analysis. However, only the logit model was mentioned in the material method section and the Wilcoxon signed-rank test was not included. After the statistical analysis comparing soft skills and the other, a discussion was made, but after the logit model, the current findings were not discussed with the previous studies. These sections should be made more regular in accordance with the flow of the article. However, another major criticism of this analysis, and therefore of the findings, is (as I mentioned earlier) that only one significance level is used in a scientific study. Therefore, either 5% or 10% significance level should be determined in Table 3.

Response 5: The authors have made changes as suggested by the Reviewer and we added the discussion to the logit model. The Wilcoxon sign-rank test is only mentioned in the methodology with reference to the relevant literature. We calculated the Z-score at level of significance 0.05; however we rewritten the table 3 more clearly as we hope.

Point 6: Instead of writing the p-value calculated in Gretl in the tables, it is sufficient to write only the p-value.

Response 6: The authors have made changes as suggested by the Reviewer

 Point 7: The abstract section can be developed with a few sentences. The sentences should be more understandable and the article should be revised in terms of language.

Response 7: The article has undergone external professional language proofreading. The abstract was added as suggested by the Reviewer

We would like to thank you for all suggestions regarding the improvement of the article quality. At the same time, we hope that it will be evaluated positively and accepted for publication in your journal.

Reviewer 3 Report

Journal: SOCIAL SCIENCES

Article title: Which skills are the most absent among university graduates in the labour market? Evidence from Slovakia

General Comments:

This article studies which skills are the most absent among university graduates in the labour market from Slovakia. The authors use a database answers by 13,684 respondents, who are university graduates of various bachelor’s- and master’s-degree programmes and completed their studies between 2008 and 2014, for 149 research activities for the project VEGA-1/0504/21, using logit model and logistic regression. The findings suggest that except the graduates in technical sciences, most graduates of the 8 Slovakian universities are not satisfied with the study programmes they chose. Moreover, the findings reveal that at the same time, employers have greater expectations in relation to soft skills than graduates have acquired. Other than foreign language skills, the level of required 11 hard skills is only slightly higher than the level acquired.

Overview:

The paper is well written and the empirical work appears to be carefully and correctly done. The research question is VERY GOOD and it does make a sufficient new contribution to the literature to be suitable for the SOCIAL SCIENCES. In fact, the literature on the satisfaction of the students with the study programmes they chose in Slovakia is quite inexistent.

The MAJOR contribution of the paper is the analysis of which skills in the particular fields of study at Slovakian universities need to be imparted better or more in order to align with the requirements of employers.

 

 

The paper is very interesting; and in my view, it needs to be MINOR improved (the methodology is somehow large, technical and need to be more applicative and easy to read) to reach the standard required for publication in this journal.

Specific Comments:

1. Introduction: good but somehow large; try to reduce it; introduce one phrase with the novelty of the study

2. Literature review: good, try to separate from introduction

3. The model: quite large and very detailed; try to reduce it; the graphs must be made into a statistical software (not in Excel)

4. Discussions: good

5. Conclusions: policy implications? Limitations?

General considerations: the idea of the article is very interesting; the results are good and with better respect for Guidelines for Contributors, Explications of the results and limitations (the introduction is somehow large, article is technical and need to be more applicative and easy to read MINOR CHANGES), it can be published in SOCIAL SCIENCES.

Author Response

Author`s responses to Reviewer`s Comments

The authors are very grateful for the reviewer`s remarks concerning the submitted article. We appreciate the comments included in the review and find them highly valuable as they indicate the path for further research within the framework of our scientific activity. Please find below our responses:

General Comments:

The paper is very interesting; and in my view, it needs to be MINOR improved (the methodology is somehow large, technical and need to be more applicative and easy to read) to reach the standard required for publication in this journal.

General considerations: the idea of the article is very interesting; the results are good and with better respect for Guidelines for Contributors, Explications of the results and limitations (the introduction is somehow large, article is technical and need to be more applicative and easy to read MINOR CHANGES), it can be published in SOCIAL SCIENCES.

Response: All general comments sa detailed described in the special comments therefore we hope that the responses to specific comments will be exhaustive to provide the responses also for this part of Reviewer`s Comments.

Specific Comments:

Point 1: Introduction: good but somehow large; try to reduce it; introduce one phrase with the novelty of the study.

Response 1: The authors have made changes as suggested by the Reviewer – we shortened the introduction and added the novelty of study which is characterised by the main objective of the paper and its partial objectives in the introduction.

Point 2: Literature review: good, try to separate from introduction

Response 2: The authors have made changes as suggested by the Reviewer – we prepared new chapter “literature review.”

Point 3: The model: quite large and very detailed; try to reduce it; the graphs must be made into statistical software (not in Excel)

Response 3: We tried to include the technical aspects of the models only in this chapter 3 (Materials and methods). We shortened the explanation of the methods used by referring to the relevant literature that deals with this problem and describes the individual procedures in detail. At the same time, we tried to reformulate the interpretations of some results as much as possible so that their relevance would not be lost and their meaning would not be distorted. Moreover, we would like to prepare the graphs in statistical software, but unfortunately our university has available only officially licensed MS Excel. We hope that the readability of the results is not limited.

 Point 4: Discussions: good

Response 4: Thank you very much.

Point 5: Conclusions: policy implications? Limitations?

Response 5: The authors have made changes as suggested by the Reviewer – we added it in the conclusions.

We would like to thank you for all suggestions regarding the improvement of the article quality. At the same time, we hope that it will be evaluated positively and accepted for publication in your journal.

 

Round 2

Reviewer 2 Report

First of all, I would like to thank the author for her explanations and corrections. The expression "*,**,and *** significant at 10%, 5%,and 1%, respectively." given as an explanation under the table in line 302 has not been corrected.

 

A scientific study should have only one level of significance . Logit model outputs should be reinterpreted according to the significance level to be selected. 

Author Response

Dear Reviewer,

we made changes as you suggested.

We would like to thank you for your suggestion to improve the quality of our paper.

 

 

Author Response File: Author Response.docx