week 5 replies

100 words each reply

D1 reply to Tammy

Descriptions, Pros, and Cons of the Four Quantitative Research Designs

In their foundational text Practical Research: Planning and Design, Paul Leedy and Jeanne Ellis Ormrod outline several primary quantitative research designs used to systematically describe and analyze phenomena. The first of these is the  Observation Study, which in quantitative research is characterized by a highly structured, respecified focus on concrete, measurable behaviors. To ensure absolute objectivity, researchers split observation periods into small segments and utilize multiple trained independent raters to count or evaluate behaviors on standardized scales. The primary advantage of an observation study is its high ecological validity, as it allows researchers to capture actual, real world behavior as it naturally unfolds, bypassing the cognitive biases of self reporting. However, its disadvantages are significant: it is incredibly time consuming, requires extensive planning and coordinate training, and remains highly vulnerable to the Hawthorne Effect, where individuals alter their natural behavior simply because they know they are being watched.

The second design is  Correlational Research, which investigates the statistical association between two or more existing variables to determine if they co-vary in a predictable manner. Researchers collect quantitative data on multiple characteristics of a single group and calculate a correlation coefficient to map the strength and direction of the relationship. The chief benefit of correlational research is its predictive utility and its capacity to identify complex patterns in situations where variables cannot be ethically, legally, or logistically manipulated. Conversely, its primary drawback is that correlation does not equal causation. Correlational research cannot prove that one variable directly caused changes in another, as it is perpetually vulnerable to unmeasured, confounding third variables that may actually be driving the observed relationship.

The third design,  Developmental Designs, focuses on tracking how specific variables or human characteristics change over the course of time. This approach is divided into cross sectional studies, which compare different age groups at a single point in time, and longitudinal studies, which follow a single group of participants over months, years, or decades. The advantage of developmental designs is their unique capability to chart growth trajectories, learning curves, and long term developmental stability. However, they carry steep operational disadvantages: longitudinal studies suffer from high participant attrition over time and can be heavily skewed by practice effects when participants take the same test repeatedly, while cross sectional studies are highly vulnerable to cohort effects, where differences between age groups are caused by unique historical or generational experiences rather than actual developmental aging.

The fourth design,  Survey Research, is a highly common quantitative approach designed to acquire structured information about a large populations opinions, demographic characteristics, attitudes, or past experiences. This is accomplished by posing a series of standardized, closed ended questions to a representative sample via written questionnaires or structured interviews, then converting the responses into numerical data. The primary advantage of survey research is its massive efficiency, enabling researchers to quickly gather easily quantifiable, generalizable data from geographically vast and diverse populations. Despite this utility, survey research is heavily compromised by response bias and social desirability bias, as participants frequently provide answers that are socially acceptable or make themselves "look good" rather than reporting their true actions or beliefs.

Which research design best fits my research topic

To determine which of these four quantitative research designs is most appropriate, the framework must be applied directly to my chosen capstone topic: Mandatory Crisis Intervention Team (CIT) Training and Police Use of Force: A Retrospective Evaluation of Street-Level Behavioral Outcomes and Citizen Complaints. When evaluating the operational realities of street level policing, Correlational Research emerges as the most viable and effective design among the four options. Because my study evaluates the relationship between the implementation of a mandatory forty hour CIT course and behavioral outcomes like physical force deployments and citizen excessive force complaints, it relies heavily on quantitative association. By examining historical, de-identified database records, this study seeks to determine if changes in the training status of a patrol force correspond predictably with a decline in high severity force incidents and administrative complaints.

Using correlational research with advanced statistical controls is infinitely superior to the other three descriptive designs for several practical reasons. First, an Observation Study is completely unfeasible for this topic; it is physically impossible, logistically dangerous, and highly unethical to place civilian research observers in patrol cars to wait around for unpredictable, high stress psychiatric crises to occur. Even if observers were present, the Hawthorne Effect would cause officers to alter their natural behavior under scrutiny, destroying the validity of the data. Second, Survey Research is deeply flawed for tracking police use of force due to severe social desirability bias. If officers are asked via a questionnaire whether training reduced their physical de-escalation tactics, they are highly likely to answer in a way that shields themselves from administrative liability or matches departmental expectations, rendering survey data highly unreliable. Third, a Developmental Design is inappropriate because my study is not trying to track the biological aging or long term lifecycle development of individual officers, but rather the immediate behavioral impact of an educational intervention.

While traditional correlational research cannot prove causation on its own, my study overcomes this textbook limitation by utilizing an advanced correlational technique: multivariate regression modeling. This approach allows the study to statistically control for critical covariates identified in existing literature, specifically suspect resistance levels and an officer's total years of service, mirroring the established methodology of previous researchers in the field (Morabito et al., 2012). By holding these confounding factors constant, the research can isolate the specific relationship between CIT training and force reduction. Furthermore, by structuring this correlational study as an interrupted time series that compares twenty four months of pre-implementation data with twenty four months of post-implementation data, the project advances past basic observation and provides a highly practical, robust, and methodologically sound evaluation of police behavioral outcomes, building upon the foundational de-escalation research frameworks established by prior scholars (Compton et al., 2014).

References

Compton, M. T., Bakeman, R., Broussard, B., Hankerson-Dyson, D., Husbands, L., Krishan, S., Stewart-Hutto, T., D’Orio, B. M., Oliva, J. R., Thompson, N. J., & Watson, A. C. (2014). The police-based Crisis Intervention Team (CIT) model: II. Effects on level of force (and arrest), referral to psychiatric services, and voluntary/involuntary receiving facility transfers. Psychiatric Services, 65(4), 523–529.

Leedy, P. D., & Ormrod, J. E. (2019). Practical research: Planning and design (12th ed.). Pearson.

Morabito, M. S., Kerr, A. N., Watson, A., Draine, J., Ottati, V., & Angell, B. (2012). Crisis Intervention Teams and people with mental illness: Exploring the factors that influence the use of force. Crime & Delinquency, 58(1), 57–77.

D1 reply to Daniela

The four types of quantitative research—observational, correlational, developmental, and survey research—each offer a different method for gathering and analyzing numerical data; the choice of design is mainly determined by the research question, the variables under investigation, and whether the researcher wishes to describe behavior, look at relationships, or find out if changes take place over time. In the case of the topic "Do Body-Worn Cameras Modify Behavior?", each of these research designs could yield useful information, but some would be more appropriate than others.

Observational research consists of systematically watching and recording behavior in accordance with previously set criteria. The main advantage of this approach is that it allows researchers to measure real behavior rather than depending completely on the participants' memories or perceptions. For instance, a researcher looking at body-worn cameras could observe interactions between police officers and citizens and note how often various behaviors occur, such as the use of force, verbal commands, compliance, resistance, or de-escalation. The researchers could also look at the footage from the body cameras and assign codes to particular behaviors. The reliability could be increased by using careful operational definitions and by having trained observers. A drawback is that observation can be very time-consuming and costly, particularly when a large number of interactions have to be coded. Moreover, observers might introduce bias, and individuals may change their behavior when they realise they are being observed.

Correlational research looks at the relationships between two or more variables without altering them. For instance, a researcher might look into whether departments or officers who use body-worn cameras more often also have different rates of complaints, arrests, use-of-force incidents, or citizen resistance. The main benefit of this kind of research is that it can detect relationships and enable researchers to make predictions; it is also useful in cases where it would be impractical or unethical to carry out an experiment. Yet correlation does not prove causation. If, for example, officers who use body-worn cameras receive fewer complaints, other factors such as department policies, officer experience, supervision, or characteristics of the community might account for the relationship. Consequently, a correlational finding on its own could not show that the cameras caused changes in behavior.

Developmental research looks at how characteristics or behaviors evolve over time and may employ cross-sectional, longitudinal, or cohort-sequential methods. For example, a cross-sectional study might compare officers or police departments that use body-worn cameras with those that don't at a single point in time. A longitudinal study would involve measuring officers' behavior before body-worn cameras are introduced and then measuring the same officers again afterward. The main benefit of longitudinal research is that it enables researchers to look at changes over time for the same individuals or groups. This is especially important in the case of body-worn cameras since behavior could change slowly as officers get used to the technology. The drawbacks include the large amount of time and resources needed, participant attrition, and possible testing or practice effects. While cross-sectional designs are cheaper and quicker, the differences between the groups might be due to pre-existing characteristics rather than to the use of cameras. Thus, longitudinal research can offer stronger evidence regarding change, although it still does not necessarily prove causation unless it is combined with an appropriate experimental or quasi-experimental design.

Survey research obtains quantitative information by posing the same questions to participants. In the case of this topic, researchers might administer a survey to police officers to find out whether they think that body-worn cameras have an effect on their behavior, on their professionalism, on the way they interact with citizens, or on their decision-making. Surveys are efficient and allow information to be gathered from large numbers of people. They are also able to measure opinions which cannot be observed directly. Yet the data provided by respondents can be influenced by difficulties with memory, social desirability, or by the respondents' readiness to give honest answers. Most important of all, the police officers' perceptions of changes in their behavior may not match their actual behavior (Ormrod, 2023).

With regard to the research question “Do Body-Worn Cameras Modify Behavior?”, an observational approach would be very useful since the dependent variable—behavior—can be directly observed. Yet, if the aim is to find out whether body-worn cameras lead to changes in behavior, a randomized experimental or quasi-experimental design would offer more solid evidence than a simple observational, correlational, developmental, or survey design. This difference is shown by the existing research on body-worn cameras. A systematic review carried out by Lum et al. (2020) looked at experimental and quasi-experimental studies dealing with outcomes such as use of force, complaints, arrests, traffic stops, and resistance; the review concluded that the effects of body-worn cameras were not consistently achieved across these outcomes.

A well-conducted study might randomly allocate similar officers, shifts, or units to a group that uses body-worn cameras and a group that does not have them. The researchers could then assess behavior both before and after the implementation by using objective measures such as reports on the use of force, complaints from citizens, arrests, instances of resistance, and the interactions that were observed. The random assignment would help to reduce the chance that any pre-existing differences between the officers account for the results. For instance, a randomized controlled trial involving 2,224 Washington, D.C. police officers compared those officers who had been randomly assigned to wear body-worn cameras with a control group and looked at outcomes such as the use of force and complaints from civilians.

In summary, observational research would be a good option for measuring the real behaviors linked to body-worn cameras, whereas a randomized or quasi-experimental longitudinal approach would be more appropriate for finding out whether the cameras lead to changes in behavior. A combination of objective observations of behavior with longitudinal measurements would offer a more thorough evaluation than depending just on surveys or correlations. This method would also overcome a significant limitation in the current literature, which has given conflicting results about whether body-worn cameras consistently alter the behavior of police officers or citizens.

Reference

Lum, C., Koper, C. S., Wilson, D. B., Stoltz, M., Goodier, M., Eggins, E., Higginson, A., & Mazerolle, L. (2020). Body-worn cameras' effects on police officers and citizen behavior: A systematic review. Campbell Systematic Reviews, 16(3), e1112. https://doi.org/10.1002/cl2.1112

Ormrod, J. E. (2023). Practical Research: Design and Process (13th ed.). Pearson Education (US). https://ccis.vitalsource.com/books/9780138088781

D2 reply to Tammy

In the text, Paul Leedy and Jeanne Ellis Ormrod outline several primary quantitative research designs used to systematically describe and analyze phenomena. A true experimental design is characterized by the researcher direct manipulation of an independent variable to observe its causal effect on a dependent variable, combined with the rigorous control of all other potentially confounding variables. The absolute hallmark of this design is the random assignment of participants or units of study to equivalent groups such as an experimental group and a control group. The primary advantage of a true experimental design is its exceptionally high internal validity, which allows researchers to confidently isolate cause and effect relationships and eliminate alternative explanations. Conversely, the major disadvantage is its lack of ecological validity or real world generalizability, as true experiments often require artificial, highly controlled laboratory settings. Furthermore, in many social science and public safety contexts, randomly assigning human subjects to potentially harmful or restrictive conditions is ethically impossible or logistically impractical.

Quasi experimental designs are utilized when the researcher seeks to determine cause and effect relationships but finds that random assignment of participants to groups is either impossible or highly impractical. Instead of randomizing, this design relies on intact, pre existing groups such as comparing two pre existing school classrooms or two different police shifts or uses repeated measures over time such as time series designs. The primary advantage of a quasi experimental design is its high practical and ecological utility, making it the premier choice for conducting evaluations in real world, operational settings like schools, hospitals, and police departments. The main disadvantage is its lower internal validity compared to true experiments because participants are not randomly assigned, the researcher cannot completely control for all confounding variables, meaning some alternative explanations for the observed changes must be carefully considered and statistically addressed.

An ex post facto design, which translates to "after the fact," is an observational approach where the researcher investigates how pre existing characteristics or prior experiences represent the independent variables and relate to subsequent behaviors or outcomes as the dependent variables. In this design, the hypothesized cause occurred long before the study commenced, and the researcher has absolutely no direct control or manipulation over the independent variable such as comparing the developmental progress of children who have already experienced historical family abuse against those who have not. The primary advantage of an ex post facto design is that it provides a highly ethical, legitimate method to study critical, real world variables such as physical trauma, inherited traits, or disease exposure where direct manipulation is strictly prohibited. The chief disadvantage is that because the researcher cannot control for confounding background variables or manipulate the independent variable, they can never definitively establish a true cause and effect relationship.

Methodological Application to the Chosen Capstone Research Topic

To evaluate the operational and behavioral outcomes of mandatory forty hour CIT training on patrol officer use of force and citizen complaints, a Quasi Experimental Design is definitively the best choice for this capstone project. A true experimental design is entirely unfeasible for this study because a police chief cannot ethically or operationally randomize use of force encounters on the street. It is impossible to randomly assign high risk, unpredictable psychiatric crises to some officers while withholding response from others in a real community. Similarly, an ex post facto design is inappropriate because CIT training is not a static, pre existing personal characteristic like gender or ethnicity; rather, it is a deliberate administrative intervention introduced at a specific point in time to alter officer behavior.

A quasi experimental framework, specifically a Simple Time Series Design perfectly matches the archival, retrospective structure of this project. By collecting twenty four months of pre implementation baseline data and comparing it to twenty four months of post implementation data, the time series design allows the study to track whether a significant change in force and complaints occurs immediately following the introduction of the mandatory training. In accordance with the principles of quasi experimental research, this longitudinal approach heavily reduces the probability that a sudden drop in use of force was caused by an outside history factor, as it would be an extreme statistical coincidence for an external event to occur at the exact moment the training was implemented.

Furthermore, this quasi experimental approach allows for the integration of multivariate statistical regression to control for key confounding variables that are otherwise difficult to manage in real world policing. By statistically controlling for suspect resistance levels and officer career longevity, the study isolates the actual behavioral impact of the CIT curriculum on street level outcomes. This analytical method mirrors advanced police evaluations in modern literature, which utilize quasi experimental and time series modeling to isolate the specific impact of de-escalation programs from broader organizational or social trends (Engel et al., 2022). By prioritizing objective, de-identified administrative records over reactive self report surveys, which are often heavily compromised by social desirability bias in training evaluations (Taheri, 2016), this design ensures a highly feasible, scientifically rigorous, and ethically sound capstone evaluation.

References

Engel, R. S., McManus, H. D., & Isaza, G. T. (2022). Evaluating the impact of de-escalation training on police behavior: Results from the Louisville Metro Police Department. Criminology & Public Policy, 21(2), 251 to 281.

Leedy, P. D., & Ormrod, J. E. (2019). Practical research: Planning and design (12th ed.). Pearson.

Taheri, S. A. (2016). Do Crisis Intervention Teams reduce arrests and improve officer safety? A systematic review and meta-analysis. Criminal Justice Policy Review, 27(1), 76 to 96.

D2 reply to Daniela

When researchers want to examine whether an independent variable is associated with changes in a dependent variable, they can use the three types—experimental, quasi-experimental, and correlational—though the amount of control the researcher has over the study differs.

In experimental studies, participants or other units of analysis are randomly assigned to either a treatment group or a control group. The researcher alters the independent variable while making every effort to keep constant other factors that could affect the outcome. The strength is that random assignment makes the groups comparable; this is why experimental designs are so valuable when a researcher wants to examine cause and effect. The main disadvantages are that such experiments can be expensive, time-consuming, and at times difficult or unethical to carry out. Moreover, results from a controlled experiment do not always generalize well to real-world situations.

Quasi-experimental research is like experimental research in that it does not include complete random assignment; rather, it involves comparing groups that are already in place or makes use of other techniques, such as looking at situations before and after an intervention or using matched groups. The advantage is that it is a practical method when it is either impossible or不合适 to carry out random assignment and it can also provide useful evidence in real-life circumstances. On the other hand, a big disadvantage is the greater likelihood of confounding variables and selection bias, which makes it harder to interpret the results. Since there is no true randomization, researchers cannot be certain of ruling out the effect of unmeasured variables which might explain the differences between the groups. For instance, if one police department uses body-worn cameras and another does not, the differences in outcomes might be due to existing departmental policies, training practices, or the sociodemographic characteristics of the communities, rather than the cameras themselves. As a result, the causal conclusions reached from quasi-experimental designs are naturally less reliable than those from randomized experiments.

Ex-post facto research deals with situations that have already occurred rather than changing the independent variable. In such research, participants are divided according to an existing characteristic, and the results are then compared. This type of study is used when it would be impossible to manipulate the variable. For example, one can compare officers who at present use body-worn cameras with those who do not. The main advantage is that it enables examination of real-world conditions that an experiment cannot create. On the other hand, the main disadvantage is limited control over confounding variables. Consequently, although ex-post facto studies can show that relationships exist, they provide only weak evidence for causation.

For the research topic "Do Body-Worn Cameras Modify Behavior?",  I would choose an experimental design. The best way to determine whether the cameras affect behavior would be to randomly assign officers to either use body-worn cameras or join a control group. The researchers could then compare outcomes such as use of force, citizen complaints, arrests, compliance, or other measures of police-citizen interactions.

Scholarly evidence already shows how this method can be put into practice. For example, Braga et al. (2018) carried out a randomized controlled trial involving more than 400 police officers in Las Vegas. They compared the officers who were given body-worn cameras with those who were not. The researchers examined several outcomes, including complaints and reports of the use of force. In another similar study, Yokum et al. (2017) conducted a randomized controlled trial with 2,224 police officers from Washington, D.C. The officers were randomly assigned to wear the cameras or serve as controls, allowing the researchers to examine behavioral outcomes under controlled conditions.

An experimental method is therefore suitable for this research question, as the researcher is especially interested in whether the presence of body-worn cameras changes behavior. At the same time, the results must be interpreted with care, since earlier studies have found different effects depending on the outcome measured and the setting. The most effective study would therefore clearly define what is meant by "behavior", establish measurable outcomes, and use random assignment where possible. In view of these methodological considerations, how could future research deal with the complexities of measuring "behavior" in various contexts or take account of the variables that randomized designs are not able to control?

References

A. A. Braga, W. H. Sousa, J. R. Coldren Jr., and D. Rodriguez (2018) 'The effects of body-worn cameras on police activity and police-citizen encounters: A randomized controlled trial', Journal of Criminal Law and Criminology, 108(3), pp. 511–538. An online scholarly source

Yokum D., Ravishankar A. and Coppock A. (2017) 'Evaluating the effects of police body-worn cameras: A randomized controlled trial'. The Lab @ DC. Online scholarly source