discussion 2

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Quantitative versus Qualitative Methods: Understanding Why Quantitative Methods are

Predominant in Criminology and Criminal Justice

George E. Higgins

University of Louisville

Abstract

The development of knowledge is important for criminology and criminal justice. Two

predominant types of methods are available for criminologists’ to use--quantitative and

qualitative methods. A debate is presently taking place in the literature as to which of these

methods is the proper method to provide knowledge in criminology and criminal justice. The

present study outlines the key issues for both methods and suggests that a criminologist’ research

questions and hypotheses should be used to determine the proper method.

Quantitative versus Qualitative Methods: Understanding the Methods in Criminology

Research is the discovery of information that is either new or replicates previous findings.

Research becomes scientific when if follows specific methodologies that others may be able to

replicate to arrive at similar results. Two types of methodologies are predominant in

criminology and criminal justice that provide this sense of science--quantitative and qualitative

methods. However, Tewksbury, DeMichele, and Miller (2005) have shown that quantitative

methods are used more often than qualitative methods in criminology and criminal justice.

Importantly, quantitative and qualitative methods differ in several ways.

The present study contributes to the literature by presenting a theoretical treatment of

quantitative and qualitative research. The study begins by presenting quantitative and qualitative

23

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

24

methods. Then, the importance of sampling to both methodologies comes. This is followed by

discussions of the primary methodologies that are used in either approach. The data that are

presented in each approach are presented. Then, the issues surrounding reliability of both

methods are presented. This is followed by the discussion.

Quantitative and Qualitative Research

Quantitative methods are based on the premise of empiricism and positivism (Rossi,

1994; Smith, 1983). These methods are rooted in the scientific method that is derived from the

physical and natural sciences. Generally, these methods allow criminologists to be objective,

formal, and systematic that arrives at a series of numbers to quantify phenomena (Creswell,

1994). That is, criminologists measure phenomena objectively affording them the opportunity to

remain distant and be independent of the phenomena that is being researched.

This is consistent with the role of values in research. Using quantitative methods,

research is able to be devoid of values. Values are removed from the research process because

statements in written reports and instruments are removed (Babbie, 2002). Criminologists argue

the “facts” of the study and not the values of the study. Criminologists that use quantitative

methods write their reports in very specific ways. First, the reports are written impersonally.

This allows them to keep their distance and to make sure that their values are not interwoven into

their research. Second, their reports are written in a formal tone with an emphasis on the

connections, comparisons, and group differences between the concepts that are being studied.

For instance, Higgins (2007) presented a report that examined the psychometrics of a specific

self-control scale. Importantly, this paper was written in a very formal tone that was removed of

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

values, but Higgins relied on the numbers to provide evidence to support or refute the hypotheses

of the study.

The issues of concepts in quantitative research are important. Criminologists use theory

to define their concepts and the connections between them. A theory is a set of interrelated or

intercorrelated concepts and propositions that are designed to explain a behavior. In

criminology, the behavior is typically criminal or deviant. Agnew (1995) argued that social

learning, self-control, and strain theories were the leading general crime theories in criminology

and criminal justice. Quantitative methods allow criminologists to be deductive in stating their

hypotheses and research questions a priori from established theory, allowing criminologists to

test theories and examine relationships for cause and effects. For example, Agnew (1992)

argued that three forms of strain generate an emotion that prepares the individual to cope with

the strain. In this example, three hypotheses are presented. The first is a direct hypothesis from

the three forms of strain generating an emotion. The second is a hypothesis that emotion

prepares an individual to cope. The third is implied and suggests that strain has an indirect

connection with coping through emotion that may be conditioned by: criminal histories, peer

association, or morality.

Qualitative methods are guided by ideas, hunches, or perspectives (Creswell, 1994;

Rossi, 1994). Criminologists that use qualitative methods are usually trying to develop theories

rather than test them. In addition, the intention is to use the language of the subject to provide

the understanding and not the quantity of the subjects. In other words, qualitative methods are

subject (i.e., study respondent) driven and not theory driven. This allows criminologists to

describe phenomena in a more humanistic and phenomenological view. Using an interview

format, the qualitative researcher would focus on coping mechanisms and then proceed

25

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

backwards as to understand why the individual coped in a certain way. This is an example of

subject generated research rather than theory driven research. The reason why the individual

copes in this manner is induced and a theory is created for understanding. Qualitative

criminologists will argue that their lack of dominance in criminology and criminal justice is due

to the belief that the development of theory is secondary or invaluable. Quantitative

criminologists recognize that falsifying the theories are far more important. To be clear, a theory

derived from 10 to 15 research subjects needs to be examined across several thousand

individuals or groups before it may be reified. This has been the case with the leading crime

theories (i.e., social learning theory, self-control theory, and General Strain theory) (Agnew,

1995). It should be noted that these theories were not developed using qualitative methods,

perhaps this is the reason why they have withstood multiple rigorous quantitative tests that

transcend disciplines, races, ethnicities, and countries.

Qualitative methods allow criminologists to become part of the study by shortening the

distance between him or herself and the research subject. Thus, they are typically the instrument

allowing them to interject his or her values into the research (Babbie, 2002). This may occur in

participant observation research where the researcher infiltrates a setting and participates in the

activity so that they may gain access and acceptability among subjects. In this form of research,

the criminologist really is the instrument and they are not able to take clear and concise notes

during the interaction leaving a substantial amount of the information to memory. Redmon

(2003) presents an example of this process. He collected interview data from individuals during

Mardi Gras; however, Redmon is unclear about how the interview data were recorded. This

leaves one to believe that he relied on his memory. If the interviews lasted between 10 minutes

and 2 hours to complete the interview how could he possibly recall of the details from the

26

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

interviews. While this is not meant to diminish the contribution that this work has made in our

field, it does illuminate the potential problems with this form of research. One of the strengths of

quantitative research is the transparency that comes from the methods that are used to arrive at

the findings.

The intention of qualitative research is not standardized and may change during the

middle of study. This occurs in content analysis and interview data. Interview data provides an

example of this issue. Maxwell (1996) argued that when performing interviews that the

researcher should use “probing” questions to gain additional information about context. The

problem with the “probing” questions is that they are often unscripted. This means that different

subjects may get different versions of the “probing” questions that may provide different

information. This is problematic when one considers that there is likely to not be an a priori

presentation of the research problem and the categories that are used to capture them.

Quantitative methods attempt to add to the universal knowledge of society (Blalock,

1979). The use of experiments, surveys, and quasi-experiments has allowed criminologists to

gain valuable insights into the criminal justice system and criminals. In criminology and

criminal justice, these methods have been used to produce “real answers” from “hard data”.

Hard data in this instance is the use of numbers. Tewksbury et al. (2005) showed that “hard

data” is the preferred method of criminologists. Qualitative methods have are generally not as

good at giving direct answers, but are good at developing more questions. This occurs because

qualitative methods are consistently using “soft data”. Soft data in this instance is language (i.e.,

body and words) to represent phenomena. Criminologists must consider sampling regardless of

the method.

27

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Sampling

Using either quantitative or qualitative methods does not absolve criminologists from the

complexity of sampling. Regardless of the methodology being used, all of the samples have to

be representative. To ensure that this takes place, quantitative methods require criminologists to

use random sampling. However, when criminologists are conducting experiments, they are

required to use random assignment (Babbie, 2002). Further, when criminologists are conducting

quasi-experiments, they are required to use some form of matching technique.

When using quantitative methods, criminologists are typically guided by the central

limits theorem to develop their samples. This theorem posits that when a sampling distribution

begins to grow it will begin to appear like a normal distribution (Blalock, 1979). This allows

criminologists to use their quantitative methods to generalize their results from their sample to

the population. Unfortunately, criminologists are not always able to achieve a random sample,

and this reduces the veracity that they can generalize their sample to the population. However,

criminologists consistently do generalize their nonrandom sample results to their populations

when they have large samples that appear large enough to satisfy the central limits theorem via

the law of large numbers.

Qualitative methods have difficulty in the area of sampling (Berg, 2007). They may use

the same strategies as quantitative researchers (i.e., random sampling: simple, systematic,

stratified, or cluster), but qualitative researchers would have to contend with large samples that

may not be specific to their research “hunch”. Qualitative researchers generally choose their

samples from individuals or entities that are germane to the “hunch” that they have that is the

impetus for the research. Generally, qualitative researchers have to contend with non-random

samples that include the following: convenience samples (i.e., available subjects), purposive

28

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

sampling (i.e., researcher uses their special knowledge and expertise about the group to select the

subjects), snowball sampling (i.e., the identification of several people with relevant

characteristics, performing the qualitative assessment, and then asking them for names of others),

and quota samples (i.e., researcher uses a matrix and then some non-random characteristic to fill

the matrix).

While some that use quantitative methods use these types of samples, the issue is that

these sampling techniques in qualitative researcher typically yield very small samples. This has

some repercussions. First, the small sample places a premium on the selective nature of the

sample. That is, because the sample is small, criminologists have to be very selective and direct

in the individuals or entities that they select to study. This means that the sample needs to have

direct importance for the criminologist’s research questions. That is, a qualitative researcher that

is interested in the lives of cross-dressing prostitutes, but only samples prostitutes from Terre

Haute, Indiana. This reduces the veracity that qualitative researchers may provide with their

theories that they covet.

Second, the small sample creates an opportunity for criminologists to influence the

results. This means that the intrusion into the environment may have caused changes that are

being processed in the results (Maxwell, 1996). In addition, the criminologists own perceptions

may shape the results given the sample is small. That is, the results from the small sample are

merely a reflection of the perceptions of the criminologist. Referring to the prostitution example,

there may only be fifteen cross-dressing prostitutes in Terre Haute. Therefore, a criminologist

that interviews five of these fifteen may not capture the full impression of what life is like in this

capacity.

29

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Third, the results from small qualitative studies are not generalizable. This means that

the “deep” and “rich” understanding that is coveted by criminologists that use qualitative

methods only applies to those or the entities that were in the study. To be clear, because the

criminologist is in the natural environment, the intensity of the research is increased and reduces

the opportunity to collect a large sample negating an opportunity to satisfy the central limits

theorem. For instance, the researcher that was interested in cross-dressing prostitution in Terre

Haute, Indiana that only captured five of the fifteen prostitutes may not be able to generalize

their results beyond these five prostitutes. That is, the lives of cross-dressing prostitutes in New

York City may be different from those in Terre Haute, Indiana.

Fourth, criminologists that use qualitative methods have to realize that their results will

be low in reliability (i.e., consistent over time). This means that their results will be less likely

be replicable because of the small sample size (Maxwell, 1996). For instance, if two

criminologists are studying the same phenomena using the same qualitative methods and

different small samples, a strong likelihood exists that the two criminologists will arrive at

different results. This may not be a function of the criminologists, but it may be a function of the

small specialized samples that they are using. Criminologists may have issues that are consistent

with methodology. In other words, the differences between the lives of cross-dressing prostitutes

in Terre Haute and New York City illuminate that the issues with reliability of the results from

qualitative research may not be in the criminologist, but the issues with reliability may be from

the small sample.

30

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Methodology

Quantitative and qualitative research requires different methods to acquire data. That is,

to address the hypotheses that come from theory, a specialized set of methods are required. A

few of these methods are: surveys, experiments, and quasi-experiments. Typically, to address

the relationship, explanatory, and descriptive hypotheses, surveys are an important quantitative

method.

A survey is a questionnaire that is designed to capture information about attitudes,

behaviors, and beliefs (Babbie, 2002). Surveys may include statements or questions to illicit this

information. To that end, the answer choices of a survey may either be open-ended or close-

ended. However, the questions or items may not be “double-barreled”. That is, the questions or

items may only address one issue at a time. Surveys allow for large distribution that makes large

samples feasible. Surveys are flexible in that they allow several different issues to be captured in

the same document. Surveys allow for standardized information to capture concepts.

Experiments are generally used for explaining causal relationships (Campbell & Stanley,

1966; Cook & Campbell, 1979). Experiments allow researchers to use well defined concepts and

topics to explain a causal connection between concepts. An experiment requires a pretest,

experimental stimulus, and a posttest. The pretest allows researchers to gain baseline

information before an experimental stimulus. The experimental stimulus is the concept that

allows researchers to determine what causes a reaction or change from the baseline. The posttest

is used to determine if the experimental stimulus created a change from the baseline. At a

minimum, experiments require the use of experimental and control groups. Experimental groups

are those that are subjected to stimulus that may create a change. Control groups are not given

the stimulus, but when they are given the stimulus they are given a placebo--a “fake” form of the

31

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

stimulus. The determination of how individuals are placed into control and experimental groups

occurs through random assignment. Experiments have the advantage because they are controlled

and isolated. Experiments are often replicable.

Quasi-experiments offer the same sort of benefits as experiments (Campbell & Stanley,

1966; Cook & Campbell, 1979). Quasi-experiments follow the same form as experiments, but

the chief difference is that quasi-experiments do not require random assignment. Typically, a

quasi-experiment requires some form of matching. When controlled properly, quasi-experiments

offer the same type of benefits. That is, they allow researchers to explain the differences in

groups. The importance of experiments and quasi-experiments has been recognized by

criminology and criminal justice that a journal--Journal of Experimental Criminology--has been

dedicated to these types of tests and studies.

Qualitative researchers rely on methods use a plethora of methodologies (Berg, 2007;

Rubin & Rubin, 1995). These methodologies include: interviews, content analyses, and

observation. Interviews are generally an interaction that my take place face-to-face, over the

phone, or through cyberspace. An interview can take place to uncover information about a short-

period of life or it may take place in order to uncover information about an individual’s life.

Content analyses are typically analyses that allow researchers to understand the importance of a

topic given its presence in certain outlets that include the media. Content analyses may be just

relevant counts of a specific form of information. The researcher would then infer about the

importance given that it only occurs a reasonable number of times. Observations may take place

in a few ways. The observer may participate in the activity or the observer may not participate.

The researcher would then try to make sense of the information that they captured.

32

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Data

The data that criminologists capture and use for analysis are different depending on the

method--quantitative or qualitative. Criminologists that employ quantitative methods capture

and use numbers. The use of numbers is important because it shows that a scientific system has

been followed (Babbie, 2002). The arrival of these numbers is important because it places this

sort of research akin to the physical and natural sciences. Thus, the use of numbers attempts to

provide criminologists with a sense of legitimacy among other scientists.

Qualitative methods rely on “soft data”. Soft data are words, body movements, or

actions. For criminologists that use qualitative methods, these are rich data that allow for a

picture to be drawn or painted about the research subject. Unfortunately, the purpose of

criminologists’ qualitative research studies is not often very clear. This leaves them in a position

where they are overwhelmed by the data that they collect. To that end, these criminologists are

often confused by the inability to limit the scope of their study. For instance, a criminologist

using qualitative interviews of ten people may have several hours of information from these

individuals. When these tapes are transcribed--transcription is a process of typing the audio

information for analysis--may be faced with twenty-five to one hundred pages of information

from one of these ten people. When the criminologist attempts to analyze the information, they

do not have the guidance from a firm research focus to guide them in making sense of their data.

Therefore, the criminologist becomes overwhelmed with the multiple themes and issues that may

be discovered in their data. Even after the criminologist is able to make sense of the data, they

have to be able to cogently present these data. For example, Pettiway’s (1996) Honey, Honey,

Miss Thang Being Black, Gay, and on the Streets is informative, but difficult to consume because

the transcripts are literally interjected into the text.

33

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Reliability

Criminologists have to be concerned with issues of reliability. Reliability is the

consistency of a result over time (Babbie, 2002). Criminologists that use quantitative methods

are more likely to generate results that are consistent over time--reliable. The reliability is a

result of the controlled environment and the standardization that may arise from standardization

in testing. For example, a survey that is administered to a random sample of 200 subjects is

likely to yield reliable results when another criminologist replicates the study.

Criminologists that use qualitative methods have to contend with this issue. Reliability is

a problem because criminologists are not able to standardize their methods (i.e., instruments)

across multiple groups. Thus, the instrument in qualitative research is often the criminologist

that is performing the observations or the interviews (Strauss, 1987). This creates an opportunity

for the assessment to be different for each individual or entity that is being examined. That is, if

two criminologists are using qualitative methods, specifically participant observation, they may

arrive at different views of the same situation. Further, two criminologists reviewing interview

transcripts may arrive at different views of the same transcripts.

Discussion

When criminologists consider quantitative versus qualitative methods, they need to

consider the strengths and weaknesses of both types of methods. This dilemma has generated the

debate over the superiority of either approach, or the type of method that should be adopted by

criminologists. Criminology and criminal justice has a strong quantitative tradition (Tewksbury

et al., 2005). However, the author resists the opportunity to promote one methodology over the

34

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

other because neither methodology has the ability to address all the questions that are present in

criminology and criminal justice.

This debate may be advantageous for criminology and criminal justice. The debate has

the ability to force criminologists to acknowledge and understand the controversies that surround

each methodology. Understanding the controversies surrounding the methods will force

criminologists to have an in-depth knowledge of the methodologies and not to be restricted to the

physical sciences. The preference of any specific methodology has implications for ideology

that go beyond that of technical strategy. That is, the choice of a method has specific

implications to the types of information that are needed; the methods that are used; and the

results that may be presented. Therefore, the choice of the methodology that is used should be

guided by the research questions and hypotheses that are being examined or developed. In

criminology, the typical focus is on testing theories. Even when researchers are evaluating

criminal justice policy, they often use theory as a guide suggesting a quantitative method.

Overall, the debate allows criminologists to create and develop better theories that will advance

criminology and criminal justice.

The debate also has the ability to illuminate quantitative research as producing more

generalizable findings than qualitative research. The generalizable findings from qualitative

research allow criminologists to make greater impacts on the criminological world then if they

relied solely on qualitative methods. That is, one reason why quantitative methods are dominant

in criminology and criminal justice research is due to the greater opportunity to generalize the

results.

Another reason that quantitative research is dominant is that anyone may be able to

perform them. This means that there is a greater opportunity for others to arrive at similar results

35

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

that are not stifled by special niches to enter certain settings. This allows for more science to

dictate the direction of criminology and criminal justice rather than “hunches” produced from

qualitative research.

References

Agnew, R. (1995). Testing the leading crime theories: An alternative strategy focusing on the

motivational processes. Journal of Research in Crime and Delinquency, 32, 363-398.

Babbie, E. (2002). The basics of social research. Belmont, CA: Wadsworth Publishing.

Berg, B. L. (2007). Qualitative research methods for social sciences. Boston, MA: Allyn &

Bacon.

Blalock, H. M. (1979). Social statistics. New York, NY: McGraw Hill.

Campbell, D. T., & Stanley, J. C. (1966). Experimental and quasi-experimental designs for

Research. In N. L. Gage (Ed.), Handbook of research in teaching. (pp. 1-76). Chicago,

IL: Rand McNally.

Creswell, J. W. (1994). Research design: Qualitative and Quantitative approaches. Thousand

Oaks, CA: Sage Publications.

Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis issues of

field settings. Chicago, IL: Rand McNally.

DiCristina, B. (1997). The quantitative emphasis in criminal justice education. Journal of

Criminal Justice Education, 8, 181-199.

Maxwell, J. A. (1996). Qualitative research design: An interactive approach. Thousand Oaks,

CA: Sage Publications.

36

Journal of Theoretical and Philosophical Criminology, Vol 1 (1) 2009

Pettiway, L. E. (1996). Honey, honey, Miss Thang being black, gay, and on the streets.

Philadelphia, PA: Temple University Press.

Redmon, D. (2003). Examining low self-control theory at mardi gras: Critiquing the general

theory of crime within the framework of normative deviance. Deviant Behavior, 24, 373-

392.

Rossi, P. (1994). The war between the Quals and Quants: Is a lasting peace possible? New

Directions for Program Evaluation, 61, 23-36.

Strauss, A. S. (1987). Qualitative analysis for social scientists. Cambridge, UK: Cambridge

University Press.

Tewksbury, R., DeMichele, M. T., & Miller, J. M. (2005). Methodological orientations of

articles appearing in criminal justice’s top journals: Who publishes what and where.

Journal of Criminal Justice Education, 16, 265-279.

37