Table of Contents
The study by Guerette (2007) used interrupted time series research design to analyze the collective effect of BSI operations. The analysis was based on the range from 1985 to 2003 which was believed to be sufficient in the determination of trends in migrant deaths. However, the analysis failed to allow ARIMA modeling. The design also involved the assessment of migrant death rates before and after implementation of the BSI agents. There was a contrast between the performance by BORSTAR rescue agents and the ordinary line agents in determining which one of them helped to save more migrant lives. Logistic regression was used in multivariate analysis of BORSTAR effectiveness (Gurette, 2007).
Type of data used for analysis
The research used varied types of data. For instance, some analysis used demographic data of the rescued or deceased migrants which is their gender, age, and their total number. Most of the data used were gathered by the BSI incident tracking system. Data from BSI Incident Tracking system was coded as 0 for rescue and 1 for death during a given incident. Logistic regression was used in the computation of ratios to compare the odds of death by each variable.
Standardized death rates per 10, 000 per month were used to compare deaths between different sectors Gurette, 2007).
Sampling procedure employed
The research specified the target population which was the migrant community. The researcher selected four Texas-Mexico sectors that were classified as the buffer zone to conclude if diffusion or displacement happened. Del Rio, Laredo, McAllen and El Paso were the buffer since most migrants detained in Tucson were moved and deported back to Mexico Gurette, 2007).
Dependent variable for the research was the migrant deaths. Data on migrant death trends was acquired from different sources as highlighted in the data section. To evaluate the effect of BSI on migrant deaths, the research relied on data from the national registration systems and some as gathered by the BSI Incident Tracking System. The range of for the inquiry used in gathering data on the dependent variable was from 1984 to 2003 (Gurette, 2007).
From the analysis of data provided by BSI Incident Tracking System, the results indicate that BSI program did not contribute to the reduction in the number of Migrant deaths. On the other hand, evaluation on BORSTAR and other rescue programs such as the Lateral Repatriation Programs that returned detained migrants from the dangerous region of Tucson to better locations along the US-Mexico border were discovered to be effective in minimizing migrant death rates (Gurette, 2007).
Bivariate analysis conducted the research based on a comparison of death rates and rescue by ordinary line agents and BORSTAR agent. The results from the bivariate analysis indicate that death to rescue ratio by BORSTAR agents is significantly superior compared to that of line agents. To elaborate on the analysis, the chances of death occurring in line agent response by line agents is 47% while the chance of migrant death when BORSTAR agents respond is 7% Gurette, 2007).
Limitations of the study
Guerette (2007) cites concern about the validity of data collected by the BSI Incident Tracking System which started recording data on BORSTAR involvement in migrant rescue in 2002. The researcher notes that the BSI system started recording about BORSTAR a year later after it started operations along the border. Also, lack of systematically recorded trends in migrant deaths restricts full understanding of BSI effect on the migrant rescue. Hence the findings from the research can only be looked at suggestively and should not be seen as facts.
Furthermore, it is difficult to determine if border patrol initiatives are the main forces in saving lives along the border considering that other humanitarian organizations carry out rescue missions for migrants in trouble.
The study by Kovandzic et.al (2004) used the multiple time series design. The design involved use of UCR data drawn from selected 188 American cities with a human population of at least 100,000 recorded from 1980 to 2000. The research examined the impact of strike laws by looking at a previous research conducted by Greenwood where a mathematical model was used to track the performance of criminals within a justice system. The results from the simulation indicate that full implementation of the law minimizes serious crimes at a rate of 28% annually at the cost of $5.5 per year.
The research design is most efficient experimental approach for assessing the effect of criminal justice policies on crime rates. The merit of the approach is that it enables the scholar to classify the three strikes law implemented cities under ‘natural experiment.” And the other 110 remaining cities in states with three strikes laws are considered to be the “treatment cities.” The final group of 78 cities is treated as the control unit for the experiment (Kovandzic et.al 2004).
Data used for analysis
The study used UCR index generated data for the period from 1980 to 2000 in 188 cities with the population of at least 100,000. The selected 188 cities had either had recorded a population of 100,000 adults or more by the start of 1990. Where; out of the 188 cities selected for the experiment,, 110 belonged to states that had implemented the three strikes legislation from 1993 to 1996.The city is considered to be the unit of analysis for the study since it is the most internally standardized and smallest unit for UCR system to generate data for a larger geographical location (Kovandzic et.al 2004).
The sampling procedure used by Kovandzic et al (2004) was determined by the multiple time series design. For instance, the city a unit was selected for use in the analysis due to its smallness and being an internally standardized unit. The study did not select states or other geographical classifications because they are vulnerable to biases and are more diverse while ignoring the vital in-state crime rates variations. For illustration, a given state could have a single jurisdiction with reasonably low crime rates for having implemented the three strikes law and harsh prison sentenced being frequently imposed (Kovandzic et.al 2004).
The study used numerous dependent variables which include; Burglary, motor vehicle theft, homicides and rape crimes committed within a city population of 100,000 people. The data used to show the crime rates trends was provided by the Uniform Crime Reports section of the Federal Bureau of Investigation (Kovandzic et.al 2004).
The study found out that where three strike laws were imposed, homicide crime rates escalated, especially for cities that had imposed the legislations by 1994. On the other hand, Cities within States that implemented the crime regulatory legislations did not witness any reasonable decline in crime.
The analysis of results presented unique findings. Crime rates in all the three groups of cities showed a declining trend in the early 1980s, and then started escalating by mid-1980s and then declined in 1990s. The pattern indicates that there were other forces behind the changing crime rates in the cities. Although all city groups witnessed a significant decline in crime rates from the 1990s, crime rates in the three strike cities decreased at a faster rate (Kovandzic et.al 2004).
Limitations of the study
The first drawback of the study is the sophistication of the research method. The multiple time series design is complicated hence any slight mistake in selecting the data for analysis leads to unreliable findings. The researchers also relied more on findings from previous researchers without considering the different context within which the other researchers were conducted. Lastly, the research did not include other significant data such as crime rates in cities without the legislation to restrict crimes.
with any paper
The study by D’Alessio et.al (1999) research made use of multiple time series research design to conclude whether the emergency cellular program helped in minimizing alcohol-related accidents on roads attended to by the program. Multiple time series design is well thought-out to be the best non-tentative approach to make casual interpretations. The research design encompasses numerous relationships between some observations done over a period and has a higher probability to be determined by a control series and an intervention not likely to be affected by a similar intrusion.
Type of data used for analysis
The study by D’Alessio et.al (1999) used longitudinal data from 1987 to 1997 (132 months). The research also relied on data (FARS). The FARS database holds all recorded data about traffic fatalities on American roads. For instance, of the 41967 traffic-related fatalities reported in 1997, 16189 were caused by alcohol abuse.
The sampling method employed in the study was grounded on the significance of the particular variable for the study. The study selected month as a unit due to the superiority associated with monthly data in comparison to yearly data when it comes to minimizing historical effects and interpretation of change over time. Also given that the study was focused on analyzing the change in alcohol-related traffic fatalities within a given state over time. Errors related to jurisdiction in reporting the fatalities were minimized D’Alessio et.al (1999).
Dependent variable of the study
D’Alessio et.al (1999) used two dependent variables which are; highway alcohol-related fatal accidents and the non-highway alcohol-related deadly accident rates. The former was operationalized by monthly crash reports by all State’s highway patrol law enforcers. All county and States highways recorded at least one vehicle being involved in a fatal crash between 8 P.M. to 8 A.M. The later consists of data recorded on all municipal roads involving a single vehicle between the same time. The idea behind using the series as control was aimed at ensuring that the emergency cellular program was not restricted to the Tennessee serviced program alone D’Alessio et.al (1999).
The findings by D’Alessio et.al (1999) indicate that there was an average decrease of 2.5% in reported fatal alcohol-related traffic accidents on both State and County roads checked by the emergency cellular program. The study also revealed no significant decline in fatal alcohol-related traffic accidents in states where the program was not implemented.
Limitations of the study
The study faces some criticism regarding the validity of the data given that some drivers fail to dial the Star THP program while on the road to report drunken motorists. Moreover, most motorists have an issue identifying their location at night which makes some of the data obtained unreliable. The ARIMA model used in data analysis raises concerns because it portrays non-significant moving-average and autoregressive parameters. Also, some of the reported accidents may have been caused by other features such as drivers using cellular phones while on the road and not alcohol.
- D’Alessio, S., Stolzenberg, L., & Terry, W. Clinton III (1999). Eyes on the Street:The Impact of Tennessee’s Emergency Cellular Telephone Program on Alcohol Related Fatal Crashes. Crime and Delinquency, 45, 4, 453- 466.
- Guerette, R. T. (2007). “Immigration Policy, Border Security, and Migrant Deaths: An Impact Evaluation of Life Saving Efforts under the Border Safety Initiative.” Criminology & Public Policy, 6,2, 201-222.
- Kovandzic, T., Sloan, J., & Vieraitis, L. (2004). Striking Out’ as Crime Reduction Policy: The Impact of ‘Three Strikes’ Laws on Crime Rates in U.S. Cities. Justice Quarterly, 21,2,207-239.