Qualitative and quantitative research methods in academic inquiry

¶ … interviews or fieldwork observations. Other research is conducted with a focus solely on quantitative data. No matter the type of study being performed, there are precise steps that should be followed by the researcher throughout every phase of the process. These steps have been honed and perfected over time so that today\'s researchers simply need to be aware of what type of study they desire to conduct and then implement the strategy specific to that type of study in order to produce the desired results. This paper will provide an overview of the types of research that can be performed, how research questions, variables and data are all related, how essential a literature review is to good research, how data is collected and analyzed, and how a questionnaire is created, piloted and validated.

Types of Research

Two types of research are available to the researcher: qualitative and quantitative research. Qualitative research focuses on methods and approaches that are delineated as naturalistic or include observation. One of the key advantages of qualitative research is that the phenomenon or subject in question can be examined comprehensively and extensively. Additionally, the data that is gathered and used in this approach is reliant on human experience and therefore deemed more convincing and potent in comparison to data collected through quantitative research. Another strength of qualitative research is that the surveys and interviews are not restricted to certain questions and can be directed and conducted by the researchers in real time (Choy, 2014). However, qualitative research is limited by the fact that its quality is largely reliant on the skills and competencies of the researcher and can be easily affected by biases and individual eccentricities of the researchers, which can in turn impact the validity and reliability of the study. The other weakness of this approach is that interpretation and examination can be time-consuming as a result of the enormous amount of data obtained (Creswell, 2013).

Quantitative research provides data that can be expressed and examined in numbers. One of the advantages of this approach is that it provides data that is descriptive. In turn, this makes it possible for the researcher to obtain a picture of the data sample or population. In particular, in this research, the correlation that exists between the independent and dependent variables are examined comprehensively. The strong suit of this is that it makes the researcher impartial regarding the findings of the study. Another advantage of qualitative research is that it necessitates a short time period for surveys that are conducted. Therefore, it is not time-consuming (Johnson and Christensen, 2008).

A limitation of quantitative research is that it lacks any kind of human standpoint or perspective. For this reason, it fails to offer a deep and comprehensive delineation of the experience of the population in the study. It lacks sufficient information, and this is a disadvantage. Another limitation is that efficacious quantitative research characteristically necessitates a large sample size, at times, numerous thousand participants, which can be just as time consuming -- if not more -- as examining and interpreting qualitative research. And just as in qualitative research, here researchers may also lack the skills and resources necessitated to conduct a comprehensive investigation (Choy, 2014).

A research problem can be delineated as a definite or clear statement regarding a field of concern, a situation to be enhanced upon, a difficulty to be eradicated, or a disconcerting question that is existent in scholarly literature, theoretically, or within a prevailing practice that points toward a necessity for significant understanding and measured investigation. It is imperative to note that a research problem does not proclaim how to undertake things, provide an ambiguous extensive proposition, or present a value question (Labaree, 2009). In research, a problem research is purposed to instigate the reader to the significance of the topic in the study. This takes into account the importance of the study, the research questions, the hypotheses as well as any suppositions made. It is also purposed to place the topic into a certain framework that delineated the factors of what is to be examined. Lastly, the research problem offers the context for reporting the outcomes and points out what is conceivably needed to conduct the study and elucidate the manner in which the finding will present this information (Bryman, 2007). Research emanates with at least one problem or statement regarding one phenomenon of interest. The research problem facilitates a researcher to emphasize thoughts, manage endeavours, and select the fitting approach, or standpoint from which to make sense of every phenomenon of interest.

Research Questions, Variables and Types of Data

There are three major types of research questions in quantitative research: descriptive, relational, and causal. Descriptive research questions focus on describing or quantifying the variables the researcher is interested in (Briggs, Coleman & Morrison, 2012). These questions often start with phrases such as \"what percentage/proportion,\" \"how much,\" \"what is,\" \"what are,\" and \"how often.\" In most cases, descriptive research questions include a single variable and one group, but may sometimes include more than one variable or group. A good example of a descriptive research question is: What is the level of achievement in Mathematics in 10th-grade students? For this question, the researcher would be interested in quantifying academic performance in Mathematics amongst 10th-grade students. This would typically involve frequencies, averages, percentages, and other measures of central tendency.

A major advantage of descriptive research questions is that they require less time and effort as far as data collection, analysis, and interpretation are concerned. In other words, the researcher only focuses on quantifying the variables(s) of interest with little or no concern for further explanation of the variables. In the above question, for instance, the researcher would only be interested in describing the number or proportion of 10th-grade students who score excellently, averagely, or poorly in mathematics -- there would be no explanation of why some students perform poorly and others excellently. This can, however, be viewed as a limitation in the sense that a descriptive research question does not lead to a comprehensive understanding of the variable(s) under study (Briggs, Coleman & Morrison, 2012). In other words, a descriptive research question cannot lead to statistical testing or verification of the research problem. This may introduce some bias.

Relational and causal research questions seek to address the limitation of a descriptive research question. A correlational research question is a research question that seeks to compare two or more groups on one or some dependent variables (Blaxter, Hughes & Tight, 2006). To make it a correlational research question, the example of the descriptive question provided above may be stated as follows: Do 10th grade students from poorer socioeconomic backgrounds achieve lower scores in Mathematics than those from richer socioeconomic backgrounds? For this question, the researcher would be interested in comparing Mathematics scores between 10th-grade students from poor socioeconomic backgrounds and their counterparts from affluent socioeconomic backgrounds.

Fundamentally, a correlational question assumes that a relationship exists between variables. This is one of the major strengths of correlational research questions -- they not only describe variables but also establish relationships between variables (Bryman, 2008). In this case, for instance, some 10th-grade students may score poorly in Mathematics due to their socioeconomic background -- due to poverty, their home environments may not have sufficient physical, psychological, and academic resources to support their learning. Nonetheless, it is important to note that association may not necessarily mean causation (Bryman, 2008). This is one of the major limitations of correlational research questions -- they do not necessarily explain causality. In this case, for instance, even if a correlation between academic scores and socioeconomic background may be found, a socioeconomic background may not necessarily be the cause of lower achievement level.

The problem of causality is addressed by causal research questions, meaning that causal research questions provide a greater understanding of the research problem compared to descriptive and research questions. In causal research questions, the researcher essentially seeks to establish causal (cause-and-effect) relationships between two or more variables in one or more groups (Burton, Brundrett & Jones, 2008). This is often achieved via experimental or quasi-experimental studies such as randomized controlled trials. An example of a causal research question would be: What is the relationship between teaching quality and achievement level in Mathematics amongst 10th students? This question indicates that achievement in Mathematics may be as a result of the quality of education as opposed to a just socio-economic background.

While causal research questions provide a better explanation of the research problem compared to descriptive and correlational research questions, a number of weaknesses cannot be ignored. First, coincidences may at times be interpreted as cause-and-effect relationships (Burton, Brundrett & Jones, 2008). Also, reaching objective conclusions can be quite difficult as the research problem may be affected by multiple variables in the social environment (Bryman, 2008). In other words, whereas causality may be established, it may not be proved with absolute certainty. For instance, academic achievement may be predicted by other factors in the social environment other than just teaching quality such as peer pressure, curriculum, and parental support.

Quantitative research questions differ from qualitative research questions. While the former seek to reach objective conclusions, the latter seek to reach subjective conclusions (Bryman, 2008). In other words, the researcher seeks to understand an individual or contextual interpretations of the research phenomenon. This premise influences the character of qualitative research questions. Instead of words like cause, influence, impact, relate, and effect, qualitative research questions are formulated using words like \"how\" or \"what.\" This implies that the researcher seeks to describe, identify, explore, or discover the research problem. These exploratory verbs inform the reader what the study will do. It is of particular importance to avoid starting a qualitative research question with the word \"why\" as this implies that researcher seeks to explain a relationship, which is a characteristic of quantitative research (Burton, Brundrett & Jones, 2008). For instance, the researcher may use an ethnographic design to explore how 10th-grade students perceive Mathematics.

Dissimilar to quantitative research questions, qualitative research questions do not have to be formulated concerning extant literature. This means that the researcher uses open-ended questions. Furthermore, the questions may change in the course of the study (Blaxter, Hughes & Tight, 2006). Indeed, in qualitative research, the researcher continually reviews and reformulates the research question.

Overall, the type and character of the research question determines the design the researcher adopts in conducting the research. It determines the variables the researcher will focus on as well as how data will be collected, analyzed, and interpreted. Proper formulation of the research question is, therefore, essential for success in research.

Role of Literature Review in the Research Process

Quantitative as well as qualitative techniques of research make use of literature reviews for explaining study elements and purpose. All effective literature reviews need to be well-ordered, presented in a clear way, and sufficiently thorough. This offers readers sufficient background information on studies on the same topic, carried out earlier. A literature review provides direction for further studies into the subject, besides ideal approaches by examining prior research works\' evaluations, findings, conclusions, and recommendations (Johnson & Christensen, n/s).

Any review of existing literature needs to take into account a number of crucial aspects. This potentially includes key facts garnered from the articles cited within one single essay or research paper\'s review. The facts need to be garnered from past researches and encapsulated within a literature review section understandable by all readers. The ideal literature review needs to elucidate the research problem under study explicitly, offer a brief explanation of the research methodology, target population, sample size, etc. Additionally, it needs to look at salient issues linked to the research. Through the literature review, researchers may also communicate the methodological and technical challenges encountered in the course of research.

A literature review proves vital in qualitative as well as quantitative researches. The latter technique involves an impartial, formal, and systematic process which makes use of figures for making and explaining findings. To elaborate, quantitative studies describe cause-effect relationships between variables. Literature review process in case of qualitative studies contributes significantly to describing and explaining research aims and outcomes (Funk, 2009). It helps provide insights into a study\'s underlying motivations, issues and views on the subject. It also aids in coming up with hypotheses or ideas for possible quantitative study.

A review of existing literature for scientific or qualitative studies aims at sharing other relevant research works\' outcomes with readers. Additionally, it serves to link to research with ongoing discussion in a subject\'s pool of related literature, bridge gaps, and extend prior research. More significantly, it offers a framework to establish the research\'s significance, besides undertaking a comparison of study outcomes with findings made by other researchers. Lastly, it aids in testing and validating theories or hypotheses, thus increasing the pool of evidence and findings and paving the way for further studies.

Data Collection and Data Analysis

Data collection and data analysis differ depending on the type of research being conducted. Experimental research, correlational studies and survey research all require large samples in order for statistical relevance to be achieved (Creswell, 2013). Qualitative research is typically more focused on a few specific cases, in which the quality of evidence (specific details about the experience or phenomenon) is provided instead of numerical/quantitative evidence. In qualitative studies, one to a few cases are needed for a sample.

Data is typically collected in these studies differently as well. Experimental research necessitates that an experiment be conducted with a control that allows variables to be tested. The correlational study like all survey research uses the survey method to collect data (surveys can be distributed and administered in any number of ways -- digitally, self-reported, in-person). Qualitative studies require data to be collected through fieldwork -- i.e., through direct observation, note taking, interview methods, and focus groups.

As each of these data collection types yield different data sets, each requires its own unique form of data analysis as well. Quantitative research typically utilizes statistical analysis (anything from T-tests to ANOVA and regression analysis). Each of these types of analysis depends on the type of study. Correlational studies for example will use correlational analysis that tests for a relationship between variables. Pearson correlation, Spearman correlation and chi-square tests are a few examples of how to analyze for studies of this type. If the study is designed to compare averages and look for differences between the means of variables, Paired T-tests, Independent T-tests, and ANOVA tests are used. Regression analysis is used when the study is designed to gauge whether one variable is a predictor of change in another variable: simple regression and multiple regression tests are suited for this type of study design (Creswell, 2013). Survey research uses descriptive analysis in which responses are described in percentages and discussed within the context of the study\'s framework. Qualitative studies\' data are analyzed using anything from content analysis to discourse, thematic or descriptive analysis. These modes will vary according to the methodology used (for example, content analysis would be suited for a systematic literature review, while thematic analysis might be suited for a phenomenological study).

Descriptive statistics and inferential statistics are used for different types of designs. For example, correlational studies will utilize descriptive statistics to measure a set of data\'s central tendency along with the way variables vary and relate to one another. A Pearson r would be a type of descriptive statistics test conducted to evaluate the strength of the relationship or if there relation goes in any one direction but descriptive statistics can also be used in causal-comparative design studies to measure data variability (Statistics for the non-statistician, n.d., p. 70). Inferential statistics on the other hand are used to compare means (typically a t-test is conducted) and statistical significance is determined by whether the p value is > or < than alpha (commonly .05) (Creswell, 2013). An ANOVA test will tell the researcher whether the differences are significant or if there are differences in the control and experimental groups (Creswell, 2013). The one-way ANOVA tests the null hypothesis using the following formula:

where µ = group mean and k = number of groups.

Examples of when these types of tests might be used can help understand the differences. Dormann et al. (2012) use descriptive analysis in their study on collinearity using Pearson r-test. Their findings show that predictor variables of r > 0.7 \"was an appropriate indicator when collinearity begins to severely distort model estimation and subsequent prediction\" (p. 1). Junco, Elavsky, and Heiberger (2012) perform inferential statistical analysis using ANOVA with students using Twitter and students not using Twitter during study. Their findings show that the group of students using Twitter \"had significantly higher difference scores\" than the group not using it (Junco et al., 2012, p. 7). The inference here is that Twitter could help students stay on course when used for the purpose of improving engagement.

Experimental Research

Correlational Study

Survey Research

Grounded Theory

Ethnography

Case Study

Type of Research

Quantitative

Quantitative

Quantitative

Qualitative

Qualitative

Mixed methods

Qualitative

Mixed methods

Sample

size

Large sample

Large sample

Large sample

One or a few cases

One or a few cases

One or a few cases

Population Type

One or more groups

One or more groups

One or more groups

One group

One group

One group

Data Collection

Experiments

Surveys

Surveys

Fieldwork

Observations

In-depth interviews

Focus groups

Fieldwork

Observations

In-depth interviews

Focus groups

Surveys

Fieldwork

Observations

In-depth interviews

Focus groups

Surveys

Data Analysis

Statistical analysis; regression analysis (simple or multiple0

Statistical analysis; correlation analysis (Pearson, Spearman, chi-square0