Crime Against Women in India

Crime Against Women in India - 8 Questions and Visualizations using rCharts, googleVis, and shiny

UPDATE: THE BLOG/SITE HAS MOVED TO GITHUB. THE NEW LINK FOR THE BLOG/SITE IS patilv.github.io and THE LINK TO THIS POST IS: http://bit.ly/1lHtVon. PLEASE UPDATE ANY BOOKMARKS YOU MAY HAVE.

Recent crimes against women, specifically the 2012 gang rape in New Delhi of a 23 year old lady, have pushed this issue as a substantially significant one for Indians to deal with. In this post, I try to address 8 different questions regarding crime against women in India. (1). How have numbers in different types of crimes and percentage of different crimes changed over the past few years for the country as a whole and for individual States/Union Territories? This is answered using data available for 12 years (2001-2012). For rest of the questions, I focus on data from 2012. (2) At the States/Union territories level, how are different types of crimes related to each other, in terms of their correlations? (3) Which States/UTs have a higher incidence of different types of crimes? (4) How can different States/UTs be classified based on the nature and extent of crime occurring in them? In question 5 through 7,  I ask questions 2 through 4 again, but for different cities.  Question 8 – Can the States, Union Territories,  and Cities be mapped along with their incidence levels of different crimes in 2012? 

Tools Used: Analysis has been done using R and visualizations have been developed using rCharts and googleVis. Interactivity has also been facilitated by the shiny server environment, thanks to RStudio. Other R libraries used include reshape2, plyr, and scales. Special thanks to Ramnath Vaidyanathan and @timelyportfolio for their super quick assistance with rCharts. Acknowledgements are due to developers of the numerous JS libraries and D3 viz folks. Extraordinary set of tools. All datasets used for this post and the code for generating charts and shiny applications are available at github.

Please Note: Since all of these charts rely on web-based technologies for appropriate rendering, the quality of the output (assuming it does) may be different depending on the computing system being used. If you are interested in static graphs, please let me know. 

Data: The National Crime Records Bureau of the Ministry of Home Affairs maintains statistics of different types of crimes occurring in the country. These generally tend to be underreported for different purposes (see here and here) but in the absence of any other source of such data, I  have used data from their website. Two primary datasets are being used. One provides annual statistics for 12 years (2001-2012) by State and Union Territory for incidence of different types of crimes.  A second dataset had to be created by converting a pdf document to a spreadsheet. These were 2012 data on the same set of crimes, but for cities. The types of crimes on which data was provided include the following 11 crimes -- Rape, KidnappingAnd Abduction, DowryDeaths, AssaultWithIntentToOutrageModesty, InsultToModesty, CrueltyByHusbandOrHisRelatives, ImportationOfGirlsFromForeignCountry, ImmoralTrafficPAct, DowryProhibitionAct, IndecentRepresentationOfWomenPAct, CommissionOfSatiPreventionAct. The variable names have been slightly edited for the purpose of the analysis. Please visit the provided links for more information on these variables.

Question 1: How have numbers in different types of crimes and percentage of different crimes changed over the past few years for the country as a whole as well as for individual States/Union Territories?

Part 1 Question 1 – For the country: The following interactive chart can help answer this question. You can interact with it in the following ways. (a) Hovering over the chart should identify values. (b) Click on the legend circles and those lines disappear/reappear. This way, you can look at the trends associated with each crime and the total number of crimes. (c)  The y-axis of the lower graph can be used as a slider to magnify (click it with your mouse and holding that position of the mouse, pull it to the right) sections of the plot.

Few observations: Total crimes and Rape have been on a rise from 2003 onwards --- 2003 had the lowest levels for these. Cruelty by husband or his relative has increased every year since 2001 and so has kidnapping and abduction (at a much higher pace). From 6851 dowry deaths in 2001, there have been over 8000 dowry deaths every year since 2007.

The next chart provides additional information related to this question. For every year, it tells us how much percent a particular crime constituted of the total crimes for that year. If you clicked on the “Expanded” version of the chart and hover over it, you should get this information. You can also click on any of the legend dots for a better perspective.


Few observations: Cruelty by husband or his relative has increased from 34.19% of total crimes in 2001 to 43.61% of total crimes in 2012, Kidnapping and Abduction has gone up from 10.18% to 15.66% from 2001 through 2012. The proportion of other crimes has remained relatively similar or has declined slightly. Of course, as the previous chart indicated, the absolute number of total crimes has gone up.

Part 2 Question 1: This interactive application let’s you visualize how different States and Union Territories (using the drop-down menu to select) have fared over the 12 years. (Don’t forget, this is just the first of 8 questions)  

Question 2. In 2012, at the States/Union territories level, how are different types of crimes related to each other, in terms of their correlations?

There were no instances of CommissionOfSatiPreventionAct.in 2012 for the country. Hence, that variable is dropped for analyses involving 2012. The correlation matrix given below provides this information. It tells you how related two crimes are to each other. The correlation coefficient, a measure of the relationship, can vary from –1 through +1. A negative coefficient indicates indicates a negative relationship (in red or a shade of red) between the two variables, whereas a positive coefficient indicates that there is a good chance of finding both crimes together.


Few observations:Rape appears to be positively related to cruelty by husband or his relative, dowry deaths, and kidnapping and abduction. Kidnapping and abduction also has a positive relationship with dowry deaths. There was only one negative relationship.

Question 3: In 2012, which States/UTs had a higher incidence of different types of crimes?

The heatmap below provides a clue to answering this question. Along the columns are different crimes (you don’t see all labels but if you hover over the rectangle tiles, you should see the crime identified) and along the rows are different States and Union Territories. Values for individual crimes are scaled to a 0 to 1 scale, thereby permitting a visualization of all variables in a better manner. Higher the value for a crime, higher the intensity of red and the scaled value will be closer to 1. Lower the value for a crime, lower the intensity of the red--- or white if crime level is negligible.


Few observations: There clearly are few states which appear to have red tiles for many crimes (e.g. West Bengal and Andhra Pradesh) but there are some states with no red tiles as well (e.g., Mizoram, Meghalaya and Manipur). We will look at this in more detail in the next question.

Question 4: In 2012, how can different States/UTs be classified based on the nature and extent of crime occurring in them?

We tackle this question using two charts. For this, I used a statistical technique called k-means cluster analysis. What this did was to classify different States and UTs into different groups (I specified a 5 group solution, in order to help interpretation, although one could theoretically have chosen a different number.) This grouping is done such that States and Union territories with similar characteristics (with respect to crime numbers) are grouped together. The first chart below tells us which States and UTs belonged to the 5 groups and the second chart following that gives information on the characteristics of these groups.


Characteristics of these groups. The chart below can be used to determine the profile of each of the 5 groups. You can select (box around a group number) and click to find few lines highlighted. This chart gives the mean incidence scores on all crime variables for members in the 5 groups. Lower the mean score in the group, lesser that crime occurring in that group. Please note that the original variable labels were truncated in this graph. The original variable labels are: Rape, KidnappingAnd Abduction, DowryDeaths, AssaultWithIntentToOutrageModesty, InsultToModesty, CrueltyByHusbandOrHisRelatives, ImportationOfGirlsFromForeignCountry, ImmoralTrafficPAct, DowryProhibitionAct, and IndecentRepresentationOfWomenPAct.


Few Observations: Group 1: (Assam, Bihar, Gujarat, Uttar Pradesh) Highest incidence of Kidnapping and Abduction and Dowry Deaths. Group 2: (Karnataka, Kerala, Madhya Pradesh, Maharashtra, Orissa) High incidence of Rape, Highest in Assault of Modesty and Importation of foreigners. Group 3: ( Andhra Pradesh, Rajasthan, West Bengal) Highest incidence of Rape, Insult to Modesty, Cruelty by husband or his relative, Immoral Traffic PAct, Dowry Prohibition Act and Indecent Representation of Women P.Act . Group 5 (Chhattisgarh, Delhi (UT), Haryana, Jammu and Kashmir, Jharkhand, Punjab and Tamil Nadu) relatively low (compared to groups 1, 2, and 3) but not the least levels (compared to group 4). Clearly, Group 4 States and UTs have the lowest incidence of crimes. These include Andaman and Nicobar Islands, Arunachal Pradesh, Chandigarh (UT), Dadra and Nagar Haveli, Daman and Diu, Goa, Himachal Pradesh, Lakshadweep, Manipur, Meghalaya, Mizoram, Nagaland, Pondicherry, Sikkim, and Tripura.

CITY LEVEL ANALYSES
Question 5. In 2012, at the Cities level, how are different types of crimes related to each other, in terms of their correlations?

As previously mentioned, the correlation matrix tells us how related two crimes are to each other. The correlation coefficient, a measure of the relationship, can vary from –1 through +1. A negative coefficient indicates indicates a negative relationship (in red or a shade of red) between the two variables, whereas a positive coefficient indicates that there is a good chance  of finding both crimes together.


Few Observations: At the cities level, Rape is positively related to Kidnapping and Abduction, Dowry deaths, and Cruelty of husband or his relatives. Dowry deaths were also positively related to kidnapping and abduction, cruelty of husband or his relatives and assault with intent to outrage modesty.

Question 6: In 2012, which Cities had a higher incidence of different types of crimes?
The heatmap below provides a clue to answering this question. Along the columns are different crimes (you don’t see all labels but if you hover over the rectangle tiles, you should see the crime identified) and along the rows are different Cities. Values for individual crimes are scaled to a 0 to 1 scale. Higher the value for a crime, higher the intensity of red and the scaled value will be closer to 1. Lower the value for a crime, lower the intensity of the red--- or white if crime level is negligible.


An Observation: Delhi and Bangalore seem to have a large number of red columns. This will be explored further in the next question.

Question 7: In 2012, how can different Cities be classified based on the nature and extent of crime occurring in them?

As we did in the context of States and UTs, we use k-means cluster analysis to classify different cities into different groups (Yet again, I specified a 5 group solution for interpretation purposes, although one could theoretically have chosen a different number.) This grouping is done such that cities with similar characteristics (with respect to crime numbers) are grouped together. The first chart below tells us which cities belonged to the 5 groups and the second chart following that gives information on the characteristics of these groups.


WOW: Delhi is in a league of its own, a lone member in a group. 
Characteristics of these groups. The chart below can be used to determine the profile of each of the 5 groups. You can select (box around a group number) and click to find few lines highlighted. This chart gives the mean incidence scores on all crime variables for members in the 5 groups. Lower the mean score in the group, lesser that crime occurring in that group. Please note that the original variable labels were truncated in this graph. The original variable labels are: Rape, KidnappingAnd Abduction, DowryDeaths, AssaultWithIntentToOutrageModesty, InsultToModesty, CrueltyByHusbandOrHisRelatives, ImportationOfGirlsFromForeignCountry, ImmoralTrafficPAct, DowryProhibitionAct, and IndecentRepresentationOfWomenPAct




Few Observations: Group 1 – Only Delhi – has the highest number of Rapes, Kidnapping and Abductions, Dowry Deaths, Assault with intent to outrage modesty, and Cruelty by husband and his relatives. Group 5, which includes Bangalore, Kolkata, and Mumbai, have relatively high incidences of assault with intent to outrage modesty and the highest levels on insult to modesty, importation of girls from foreign country, immoral traffic P. act, and dowry prohibition act. Group 3, which includes Ahmedabad, Hyderabad, Jaipur, and Vijayawada, has relatively higher number of insult to modesty and cruelty by husband and his relatives. It has the highest incidents with respect to indecent representation of women p.act. Group 2 – has the fewest number of crimes among all groups, and therefore likely to be the safest in this set of cities. These include cities such as Allahabad, Amritsar, Aurangabad and Chandigarh (for a complete list, please see the previous chart). After Group 2, Group 4 appears to be the safest. This group includes cities like Agra, Asansol, Bhopal, and Chennai (for a complete list, please see the previous chart).

Question 8: Can the States, Union Territories, and Cities be mapped along with their incidence levels of different crimes? Yes. Please interact with the application below and toggle between the two tabs for information regarding Cities or States and Union Territories.