GIS-Based Statistical Analysis of Detecting Fear of Crime with Digital Sketch Maps: A Hungarian Multicity Study
Round 1
Reviewer 1 Report
The authors evaluated fear of crime perception and official crime statistics in a spatial context, by employing digital sketch maps and statistical GIS methods in Hungarian cities. In sum, I liked the approach and the paper is well structured. However, it seems there some points in the manuscript that needs improvement. I tried to show them with some comments below. My overall recommendation is to reconsider after revision. Therefore, if the author addresses everything I pointed out, it is gonna be fine.
1- In my opinion, it was a mistake to aggregating different crime types (“…we applied only the categories of property crimes and crimes against people…”). A pickpocket generates a different impact in comparison to a killer in terms of fear in a location. Please, take a look into the articles:
Haberman, C. P. (2017). Overlapping hot spots? Examination of the spatial heterogeneity of hot spots of different crime types. Criminology & Public Policy, 16(2), 633-660.
Melo, S. N., Matias, L. F., & Andresen, M. A. (2015). Crime concentrations and similarities in spatial crime patterns in a Brazilian context. Applied Geography, 62, 314-324.
Briefly, they stated that aggregating crime types in a study that implicitly or explicitly considers the spatial patterns of crime is improper. A straightforward suggestion: just mention that you are aware of this limitation.
2- Figure 5 – (a) spatial distribution of crime events OR spatial distribution of KDE crime events?
3- Where are the OLS regression diagnostics? Please inform it to the readers.
4- In the conclusion section, I suggest the authors add a paragraph describing the study limitations and making recommendations to future researches.
Author Response
Cover letter
for revisions of the original manuscript entitled
GIS-based Statistical Analysis of Detecting Fear of Crime with Digital Sketch Maps: a Hungarian Multicity Study
by Ákos Jakobi and Andrea Pődör
submitted for publication at International Journal of Geo-Information
(Manuscript ID: ijgi-749925)
We are very grateful for the comments of the reviewer! The correction of the manuscript was done as follows (for easier follow up, we copy the reviewers’ comments in this letter):
Revisions made according to comments of reviewer 1:
“1. In my opinion, it was a mistake to aggregating different crime types (“…we applied only the categories of property crimes and crimes against people…”). A pickpocket generates a different impact in comparison to a killer in terms of fear in a location. Please, take a look into the articles:
Haberman, C. P. (2017). Overlapping hot spots? Examination of the spatial heterogeneity of hot spots of different crime types. Criminology & Public Policy, 16(2), 633-660.
Melo, S. N., Matias, L. F., & Andresen, M. A. (2015). Crime concentrations and similarities in spatial crime patterns in a Brazilian context. Applied Geography, 62, 314-324.
Briefly, they stated that aggregating crime types in a study that implicitly or explicitly considers the spatial patterns of crime is improper. A straightforward suggestion: just mention that you are aware of this limitation.”
Line 156-159: We found the reviewer’s related comment highly valuable. The text was supplemented by the awareness of crime type-related differences and the suggested references were also included both in the body text and in the list of references.
Line 191, 268, 304, 313: inclusion of new references resulted changes in reference numbering
- Figure 5 – (a) spatial distribution of crime events OR spatial distribution of KDE crime events?
Line 277-279: The caption of Figure 5 was corrected to clarify that maps have been created by the application of KDE crime event data.
Line 190: In addition to the above mentioned changes, the abbreviation of KDE (kernel density estimation) was described more precisely in the text.
- Where are the OLS regression diagnostics? Please inform it to the readers.
Line 372 and 379: (Table 3) The results table of OLS regressions was supplemented by a new row of VIF (max) referring to the highest Variance Inflation Factor for testing multicollinearity. Since VIF values for all independent variables in a multivariable model is under this highest value, one can decide the existence or ignorance of multicollinearity (hence providing VIF values for all involved variables is unnecessary). Here, all VIF statistics happened to be low, therefore no problematic multicollinearity was observed among the applied variables.
Line 392-395: readers were informed about VIF results of OLS regression diagnostics in the text.
- In the conclusion section, I suggest the authors add a paragraph describing the study limitations and making recommendations to future researches.
Line 417-423: Study limitations, such as sample size related significance, lack of direct explanation of markings or limitations of measuring the feeling of fear is shortly discussed in the chapter of conclusions.
Line 428-429: Although recommendations for further studies have already been included in the paper, the text was supplemented by an important future possibility: the application of spatial lag models.
Thank you for your consideration!
Sincerely,
Ákos Jakobi and Andrea Pődör
Reviewer 2 Report
The topic is interesting both the collection of data and comparison of data with official crime statistics. The benefit is that the comparison is processed for nine cities (formerly for 18 cities they have digital sketch maps) not only for one city.
Some improvements and clarifications will be beneficial:
Row 149- 155: Description of type of crimes. Please describe the types of crimes in more detail that was considered. If it was: murder, rape, physical assault, theft from cars or crime connecting traffics (car accidents) etc. I assume that "white collar crimes" was excluded. From the point of feeling, sometimes dangerous crossroads are considered as unsafe places. Maybe discus it in part 4. Discussion.
Row 196 - Add to the title of chapter "3.1 Spatial patterns analysis of people opinions/feeling" to better express the topic of analysis.
Figure 2 a) are safe and b) is unsafe. The color scale will be better in green colors scale for a) according to source color for data collection. And red color scale for all b) unsafe. It will be better to have two color scheme=two legends, not only one for both columns of maps.
Row 289-311 is one long paragraph. There is description of Table 1. Please, divide it to 4 paragraphs according the described cities.
At row 294 start second paragraph, row 304 third paragraph and row 308 last paragraph.
Table 1 - explain that N is number of cells.
Author Response
Cover letter
for revisions of the original manuscript entitled
GIS-based Statistical Analysis of Detecting Fear of Crime with Digital Sketch Maps: a Hungarian Multicity Study
by Ákos Jakobi and Andrea Pődör
submitted for publication at International Journal of Geo-Information
(Manuscript ID: ijgi-749925)
We are very grateful for the comments of the reviewer! The correction of the manuscript was done as follows (for easier follow up, we copy the reviewers’ comments in this letter):
Revisions made according to comments of reviewer 2:
- Row 149- 155: Description of type of crimes. Please describe the types of crimes in more detail that was considered. If it was: murder, rape, physical assault, theft from cars or crime connecting traffics (car accidents) etc. I assume that "white collar crimes" was excluded. From the point of feeling, sometimes dangerous crossroads are considered as unsafe places. Maybe discus it in part 4. Discussion.
Line 152-156: A more detailed view of the excluded and included crime types are provided in the text.
Line 419-423: One could have a feeling of fear because of many reasons (including also traffic-related issues), which can only be directly investigated through detailed questionnaires. This limitation of the study is expressed in the text.
- Row 196 - Add to the title of chapter "3.1 Spatial patterns analysis of people opinions/feeling" to better express the topic of analysis.
Line 202: The chapter title was modified accordingly: “Spatial pattern analysis of the opinions of people”.
- Figure 2 a) are safe and b) is unsafe. The color scale will be better in green colors scale for a) according to source color for data collection. And red color scale for all b) unsafe. It will be better to have two color scheme=two legends, not only one for both columns of maps.
Line 213: (Figure 2) Colors and legends have been modified to better follow colors applied in Figure 1. Shades of green for safe and red for unsafe places were used. The darker the color, the more times the area is marked.
- Row 289-311 is one long paragraph. There is description of Table 1. Please, divide it to 4 paragraphs according the described cities. At row 294 start second paragraph, row 304 third paragraph and row 308 last paragraph.
Line 299-325: The originally long paragraph was divided into 4 paragraphs to better separate different city results.
- Table 1 - explain that N is number of cells.
Line 297: (Table 1) Footnote was complemented with the exact description of “N”.
Thank you for your consideration!
Sincerely,
Ákos Jakobi and Andrea Pődör
Round 2
Reviewer 1 Report
I accept the paper in the present form.
