Houston is the largest city in Texas, world-known for its multi-culture and energy industry. That might be the reason why there is a huge influx of population immigrated there. Such a multi-ethnic population structure not only stimulates a rapid growth of economy in Houston, but also gives rise to some serious social problems.
Among diverse social problems, I focus most on the problem of violent crime especially homicide. Partly because violent crime especially homicide is the most brutal crime to human life, which will lead to a bad impact on public stability in society. That is why I conduct this study to examine a relatively bigger image of homicides in Houston.
1.2 Data Information
To start a study, we need to define the populations we are interested, sample from them and get a set of good data. The data “homicides in Houston from 1985 to 1994” we use in this research has been well collected by FBI. It contains 58 types of variables such as gang-related homicides, drug-related homicides, victims and offenders’ ages, victims and offenders’ genders, methods of killing and so on. All those variables are the characteristics of our sample or population interested and each of them has its series of data values.
1.3 Statistical Techniques
Using SPSS software, we will apply some basic techniques of statistics in this paper to: 1) describe the personal characteristics of victims and offenders (gender, age and race) with descriptive statistics (central tendency, dispersion and graphs); 2) construct confidence intervals around a sample proportion of drug-related homicides to estimate its population proportion; 3) conduct a Chi-Square test to examine the association between gender of offender and method of killing; 4) conduct a T-test to see the difference between the average age of the female offenders and that of the male offenders; 5) have a correlation and regression analysis for the relationship between the Gini coefficient variable and the homicide rate variable.
During all the statistical processes, we will describe and interpret the results in detail to make some inferences about the population characteristic of homicides in Houston. And in the end of the paper, we will try to make a conclusion about all the findings above with a hope to help the criminal justice department of Houston to better solve the crime problems in the future.
II. Means, Medians, Modes and Standard Deviation of 3 or More Variables 2.1 Variables of the Gender, Age and Race of Victims
| gender of victim| victim age category| race of victim| N| Valid| 4944| 4655| 4832|
| Missing| 0| 289| 112|
Mean| .8265| 3.6324| 2.1502|
Median| 1.0000| 3.0000| 2.0000|
Mode| 1.00| 3.00| 2.00|
Std. Deviation| .37875| 1.15216| .76658|
Sum| 4086.00| 16909.00| 10390.00|
Table 1: Descriptive statistics of victims’ gender, age and race
Table1 shows us some personal characteristics of victims including gender, age and race. From the first column, according to the valid N with 0 cases missing, we can see that there are in all 4944 people murdered in Houston from 1985 to 1994. Because gender is a nominal-level variable, the codes we assign to each type of this variable (male=1, female=0) has no numerical meaning. Then the standard deviation ( .37875) is meaningless. The mean ( .8265> .5) and the median (1) tell us that there are more male victims killed than female victims. The mode (1) tells us that the most frequent gender of victims is male, in other words, there are more male victims being killed. The sum is 4086, and it means that there are altogether 4086 male victims killed.
From the second column, we can find that the total number of victims is also 4944 (4655+289) in which 289 are missing due to either subject-failure or system-failure during the research. From the mean (3.63≈4), we...