Analysis of Research Methods and Designs

Introduction

Research method refers to the approach to inquiry which employs empirical scientific approaches. Conducting research involves using the scientific method at its core. Therefore, before any research is undertaken, it is important to be aware of the steps.

Research methodology has not come up overnight, but has evolved through hundreds of years of science. The history of science is interesting and intriguing, giving an insight into the developments of modern-day science. It encompasses the methods of data collection, methods of analysis, and the research designs which flows coherently from the epistemological, ontological, and philosophical world-views. Osuala (1993) argued that research is the process of arriving at dependable and objective solutions to problems through systematic data collection, analysis, and interpretation. Research aims at discovering the relationship between phenomena. Creswell (2009) held that research design incorporates plans and steps taken to finding answers to research question. It involves strategies, data collection methods, analysis, and interpretation. Creswell (2009) posited that the world-view assumption informs the decision of choice of design. According to Creswell (2009) world-view is “a basic set of the belief which guides action and methods of investigation’’ (p. 6). Creswell (2009) identified four world-views which shape the choice of research methods and design. They are the post-positivism, constructivism, advocacy, and pragmatism.

Research strategies are the specific design or models within the qualitative, quantitative or mixed methods of research. They are equally called the approaches to inquiry (Creswell 2009). Research methods refer to the forms of data collection, analysis and interpretation employed by researchers.

Methods and Design for Descriptive Research Analysis

Descriptive research enables the investigator to specify the nature of given phenomena. According to Osuala (1993) descriptive research is a method of inquiry which gives the picture of the population. As the researcher seeks to understand and describe the population, the variable units of the population need to be studied carefully as the basis of analysis. Osuala (1993) argued that the importance of descriptive research flows from the need to gain a full understanding of the phenomena. The need to make important policy decisions makes accurate description of the phenomenon imperative. Osuala (1993) however, held that there is a need to make a clear distinction between policy decisions and outcome of scientific research.

Descriptive research may provide the information needed for policy formulation, but it must be clearly distinguished from the actual policy. According to Osuala (1993), policies are the ultimate result of value decisions deduced from the research, but the policies cannot be said to be determined scientifically. According to Osuala (1993) descriptive research is often confused with decision making and the basis for policy formulation. Osuala (1993) argued that the result of descriptive research does not account for other variables which may predict the behavior of the phenomenon of study, as such cannot be generalized. He held that though descriptive research is a prerequisite for finding answers to question, it is not adequate in itself to provide answers to questions as it cannot establish cause and effect relationship.

The failure of descriptive research to provide tools for generalizing results is a great demerit of this research method as means of gaining knowledge. Osuala (1993) cautioned social scientists and educational investigators from assuming full answers in the description. According to him the result of descriptive research cannot empower the investigator to deduce the cause of the phenomena or predict what the future phenomenon will be. Despite the glaring shortcoming of descriptive research, it is a valid means of scientific inquiry as it enables the researcher gain in-depth understanding and description of phenomena. Osuala (1993) argued that descriptive research could use the same design done at different times and specified periods, show the trends in the description from which hypothesis could be formed, and later tested under controlled experimental conditions. Where no systematic knowledge is available about the population, descriptive research is imperative for gaining initial understanding of the phenomenon.

The traditional case study of an individual, a group of individuals, a school, a community, or any other unit characterized by efforts to learn as much as possible about the population or phenomena is a common design of descriptive research. According to Osuala (1993) the absence of systematic knowledge about the phenomenon is the key motivation for descriptive research. Osuala (1993) posited that longitudinal studies aim at generating hypothesis for experiments. Longitudinal description involves seeking to understand societal affairs over time through systematic review of trends, discontinuities, and shifts in emphasis. Another aspect of descriptive research, according to Osuala (1993) is selected description where the interest of study is certain aspects of a population or phenomena. Descriptive research design allows for mirrored study of the targeted aspects of the population. This is most evident in a qualitative study where information-rich samples are selected purposefully for study, Creswell (2009).

The common statistics for descriptive research are frequency distribution, percentiles, percentile rank, averages and standard scores. Osuala (1993) posited, “Descriptive statistics provide measures for reducing raw data to easy-to-understand dimension” (p.149). According to Osuala (1993), descriptive statistic provides useful descriptive information about data, which enables the investigator answer the needed research question to understand the phenomenon. Osuala (1993) held that methods of securing data for descriptive research are direct observation, indirect observation and analysis of existing records.

Observation enables the researcher to study the phenomenon by direct involvement in the process as a participant-observer or non-participant observer. An observer is able to watch the phenomenon in it’s natural-unfolding form, study and describe it. Indirect observation is akin to using interview or using an observer to describe the result of his observation. This method is referred to as interview by Creswell (2007). Analysis of record is a secondary source of data collection where the inquirer resorts to analyzing records of trends or phenomenon.

Empirical Review of Descriptive Research Methods

The choice of research method largely depends on the philosophical world view which best provides blueprint for answering the research question. Creswell (2009) argued that philosophical assumption underscores the research approach, the strategies, methods, and design. Osuala (1993) did not relate the descriptive research method with any known philosophical worldviews. There is a disharmony between the caption used by Osuala (1993) and the content of the discussion. Osuala (1993) posited that descriptive research only aims at informing the right understanding of the variables, population, or the phenomenon of interest. It is therefore, not a method but an analytical approach at determining the phenomenon. Osuala (1993) held that three basic methods of data collection in descriptive research are direct observation, indirect observation, and documents analysis. According to him indirect observation is when the researcher seeks to secure data from those who saw. The explanation revealed a redundancy in word usage here as secondary information gotten through either interview or survey questionnaire does not amount to observation whatsoever.

Descriptive research by itself cannot answer a research question bothering on the assumption of relations between variables as it is not sufficiently structured to test such relationship. The purpose for which a researcher would embark on a longitudinal description is unclear. Such a design is time consuming, costly, and redundant in purpose as according to Osuala (1993) descriptive research only aims at understanding the variables of the study. Osuala (1993) admitted that generalization is beyond the province of descriptive research and that many social scientist and investigators have been misled by relying on the result of descriptive research to assume full answer to the research questions. The method is therefore, not recommended for a scholarly empirical inquiry.

Experimental Research Method

Experimental research method enables the research study relationship between randomly assigned variables in a controlled environment. The more randomized the experiment, the better ability to generalize findings to the population of the study. According to Osuala, (1993) the purpose of experimentation is to derive verified functional relationship among phenomenon under controlled conditions. Experimentation is a method of investigation applied by philosophical scholars in a quantitative strategy of inquiry. According to Creswell (2009), experimentation aim at investigating the outcomes from the effect of treatments or manipulations. True experiment exhibits the highest degree of validity and abilities for generalization.

Experimental design is the ideal strategies for quantitative investigators. Osuala (1993) argued that experimentation is the most scientifically sophisticated research method. The genesis of experimentation, according to Osuala (1993) began with Galileo’s experiments in 1589 which demonstrated that bodies of the same substance fall at identical rates of speed regardless of their mass and of Pasteur’s discovery that food spoilage can be attributed to bacteria. Osuala (1993) posited that application of experimentation in the social sciences was limited by the complexities of the phenomenal relationship until the development of multivariate analysis. Human behavior and actions are shaped by the combination of factors, as such attempts to explain or predict it by the manipulation of causal variable alone provides inadequate explanation. The development of multivariate analysis however, allowed for the study of extraneous factors mediating the relationship between the variables.

Experimentation simply enables the researcher to improve the conditions under which he/she observes and, thus, to arrive at a more precise outcome. An experimental researcher begins by selecting and defining the problem of study, review relevant literature, state the research design, identify the population and sample of the study, state the sampling strategy, conduct the study, assess the outcome, analyze and interprets the result, write the conclusion and communicate the result. Control in experimental design involves manipulating the independent variables with the view to understanding the degree to which it predicts the state of the dependent variable. According to Osuala (1993) the strength of experimentation stems from it’s testable validity and objectivity.

Experimental design however, suffers some practical limitations. The drawback of experimentation Osuala (1993) held, is its complexity. He therefore, advised researchers to conduct a pilot study before embarking on full experimentation. The larger the sample size, the greater the chances that they characteristically predict the population and therefore, the more valid the generalizability. This position is opposed to qualitative method where the smaller the sample size, the easier it is to target information-rich cases for in-depth understanding Creswell (2007). An experimental design requires close adherence to plans with respect to factors of control, randomization, and replication. Osuala (1993) argued that the duration of the experiment should be such that the variable is given adequate time to provide changes that can be evaluated and to insulate the influence of such extraneous factors as novelty. Herein lies the confusion between experimental design in social and natural sciences.

Medical laboratory investigation where bacteria is isolated and allowed to grow under a culture and sensitivity environment is a scientific experiment. Osuala (1993) however, posited that experimentation in social science is more complex than it is in medicine. He argued that the testing of drugs and vaccines is based on a simple experimental design. The complexities of social science experiment are ameliorated by the development of multivariate analysis. The advent of computer analytical software such as Statistical packs for Social sciences has made social science experiment friendlier. Osuala (1993) stated that data analysis in experimental design is challenging and complicated. He argued that the researcher must be sensitive to the possibilities that the results of his study arose through the operation of uncontrolled extraneous factors. He/she must therefore, insulate at a given probability level, the possibility that his experimental findings are simply the result of chance.

The deployment of statistical procedures in the analysis of experimental data is highly imperative as the basis for valid interpretation. Osuala (1993) argued that in no other area of research is the need for competence in statistical procedure so prominent than in experimental design. He held that statistical tools do not relieve the investigator of his responsibility for planning the study, for controlling extraneous factors, and for obtaining valid and precise measurement.

The aim of experimental design is to be able to refer that a given independent variable ‘X’ predicted the change in the dependent variable ‘Y’. In other to make such interference with a degree of certainty, the researcher will control other possible variables impacting on the dependent variable. Osuala (1999) identified such variables constituting potential threats to internal validity of experimental design as interaction of selection and maturation, experimental mortality, statistical regression, measuring instruments, pretesting procedure, maturation processes, contemporary history, and differential selection of subjects. Potential threats to external validity according to Osuala (1993) are interaction effects of selection bias and independent variable, reactive or interaction effects of pretesting, reactive effects of experimental procedures, and multiple treatment interference. Internal and external validity enable the research to make valid inference and generalize findings with a degree of certainties.

Empirical Review on Experimental Research Method

Experimentation is a rigorous research design in which variables are randomly assigned to study groups in other to determine the effects or outcomes among the same. Osuala (1993) offered a generous historical background of experimental design. The key attributes of experimentation are randomization and control. Once these are achieved the ability to generalize is secured. Osuala (1993) however, failed to appraise experimentation in its relation to enhancing the internal validity of the research as opposed to other designs.

Factors which limit the use of the experiment in quantitative design were not discussed. Osuala (1993) did not inform readers that certain social attributes of human character such as race and gender cannot be controlled and randomized thereby limiting the use of experimental design by investigators. According to Creswell (2009) experimentation involves administering treatment to one group while withholding it from the other, to measure how the two groups score to find the difference and determine the effect of the treatment. This raises a lot of ethical issues when human participants are used. Osuala (1993) did not disclose the ethical challenges and other limiting factors to the use of experimentation in certain instances.

Though the study did not highlight the shortfall of experimentation, Osuala (1993) was able to show how researchers could overcome the inherent limitation of using experimental design to predict human behaviors. According to him the advent of multivariate analysis to account for the multiple predictors made the application of experimentation to human behavior a scientific possibility. Osuala (1993) detailed steps to be taken in conducting experimental research. The demonstration by Osuala (1993) exhaustively explained the concept of experimentation in scientific research. He however, did not align experimental design to any of the three main research approaches as shown by Creswell (2009). He did not discuss the epistemological alignment of the strategy to any world views. The study is informative and scholarly beneficial.

Data Collection Methods

Data collection method is the process of identifying and obtaining raw facts and figures relating to the phenomenon of study. Bad data will bring out misleading results of the research endeavor. Frankfort-Nachmias and Nachmias (2008) listed four methods of data collection as direct observation, survey research, secondary data analysis, and qualitative research. They also discussed the concept of triangulation which enables data to be collected using multiple strategies.

Observation

Observation as a method of data collection enables the investigator to personally witness events as it unfolds and record the findings for further analysis. Frankfort-Nachmias and Nachmias (2008) stated that observation aids the researcher to watch events, understand, explain, and predict a phenomenon. The observer must pre-determine what to observe, where and when to observe, how to record the data, and how much to infer when recording observation. It means that the observer must not be distracted with other observable interests which do not relate to the phenomenon. The greatest strength of observation is that it enables the researcher to study the behavior as it occurs directly without being subjected to transcription by other secondary party. According to Frankfort-Nachmias and Nachmias (2008) the process of transcription is often subject to attendant risk of distortion. They stated that observation has the advantage flowing from its directness. During observation the researcher collects data as such minimizes the chances of introducing elements of artificiality into the research environment.

The best form of observation is done when participants are not aware that their behaviors are being observed or when they had become familiar with the observer. This however, poses challenges of ethical consideration. According to Frankfort-Nachmias and Nachmias (2008) observation is indispensable as data collection method when a researcher is studying participants who cannot express themselves introspectively, unable to verbalize or exhibits attentive concentration on a task required. A silent observer catches the participant in their natural state. This is so because when people are aware their behaviors are being observed, they tend to modify or change such behaviors and present a distorted picture of the behaviors under study.

A silent observer does not interrupt the events being observed. Frankfort-Nachmias and Nachmias (2008) argued that a simple researcher has no responsibility for structuring the situation being observed thereby eliminating bias from the process. They argued that simple observation ensures that data are collected without interruption to the process or function of the people being observed. Observation therefore, does not obstruct the event or behavior being studied. The most pronounced limitation of observation however, is that it may infringe upon ethical considerations.

A silent observer infringes on the rights of participants. A silent observer may indeed unintentionally and uncontrollably cause changes in the events during the observation thereby introducing bias to the process. Frankfort-Nachmias and Nachmias (2008) held that the sample being observed may be so confined that they do not adequately represent the population thereby impeding the ability for generalizations. The disadvantages of observations notwithstanding, Frankfort-Nachmias and Nachmias (2008) posited that it enhances the validity of findings as it enables researchers employ triangulation through the use of two or more methods of observation. They classified observations to controlled and uncontrolled types.

Controlled observation is when, what, how, and where to observe are made clear and explicit whereas, uncontrolled observation is more flexible and less systematic. According to Frankfort-Nachmias and Nachmias (2008) research problem and design drive the choice between controlled and uncontrolled observation. They aligned controlled observation mostly with experimental researchers.

Survey Research

Survey is a method of data collection which enables the researcher to obtain data from sources not amenable to employ observation. A survey is used to garner attitudes, opinions, and describe trends of a population by selecting and studying representative samples, questionnaire or structured interviews to collect data used for generalization from the sample to the population. Creswell (2009) argued that survey provides a quantitative or numeric description of trends, attitudes or opinions. According to Frankfort-Nachmias and Nachmias (2008) certain phenomenon are not directly accessible by researchers and so data can only be collected from those who experienced the phenomenon. The responses of those who experienced the phenomenon represent the data which are used to test the stated hypothesis or answer the research question. Frankfort-Nachmias and Nachmias (2008) argued that there are three major types of survey which are the mail questionnaire, personal interview, and telephone interview

Mail Survey Questionnaire

Mail questionnaire is cheap to administer to large and widely dispersed respondents. According to Frankfort-Nachmias and Nachmias (2008) mail survey ensures the anonymity of respondents and minimizes bias hence interactions between the investigators and the participants are reduced. Participants’ protection is a crucial issue in the institutional and ethical requirement of social science research. Mail questionnaire ensures that the respondent is confidentially veiled thereby assuring anonymity particularly in sensitive and embarrassing issues that may prejudice them. The wide coverage of mail questionnaire makes it a better choice in a quantitative design where large sample size bestows more ability for generalization of the findings to the larger population. Frankfort-Nachmias and Nachmias (2008) posited that mail questionnaire is a preferred choice when participants need time to reason out their responses. In a study involving large samples, the cost of interview could be prohibitive as such mail questionnaire is the most cost-effective method of data collection.

The advantages of mail questionnaire notwithstanding, there are glaring drawbacks to its use as a data collection strategy. Frankfort-Nachmias and Nachmias (2008) argued that the wordings of the mail questionnaire pose a lot of challenges to researchers, as ambiguity could result in misunderstanding by the respondents. As opposed to interviewing method, the mail questionnaire does not afford the investigator the opportunity to probe beyond the answers given by the respondent. Questionnaire in a quantitative research are more structured and close-ended than the interview method which is more popular with qualitative open-ended and inductive. There is no guarantee that the targeted respondents actually fill out the questionnaire as some may delegate the job to subordinates or friends who may not be grounded in the phenomenon of study.

The researcher using the mail questionnaire is not in control of the actual delivery of the mail questionnaire. Frankfort-Nachmias and Nachmias (2008) argued that unintended persons therefore, may end up completing the questionnaire. This inputs error of sample selection and will impair the validity of the study. According to Frankfort-Nachmias and Nachmias (2008) a serious limitation of mail questionnaire is low response rate, as opposed to a personal interview which has a higher response rate. The low response rate is due to failure of a sizeable number of respondents in completing and returning the questionnaire. Low response rate could seriously impede the ability to generalize the findings of the study. The low response rate is associated with the literacy levels of the respondents. Questionnaire generally is not amenable to illiterate respondents who may need the services of an interpreter thereby increasing the chances of transcription error. The involvement of an interpreter could defeat the advantage of anonymity associated with questionnaire.

Personal Interview

Interview is a face-to-face and interpersonal method of data collection. Frankfort-Nachmias and Nachmias (2008) held that personal interview is an interactive process where the researcher asks the respondents’ series of structured questions aimed at eliciting responses relative to the research hypothesis. The questions, the wording, and the order in which they are asked are driven towards answering the research questions. Frankfort-Nachmias and Nachmias (2008) identified three types of personal interview as the schedule structured interview, the focused interview, and non-directive interview. The schedule structured interview employs identical question all through and follows the sequence in which the questions are asked. The non-directive interviews according to Frankfort-Nachmias and Nachmias (2008) is the most flexible form of interviewing. There is no predefined schedule of questions but the respondents are given the opportunity to recant their experiences and with utmost independence.

Personal interview is more aligned to a qualitative inquiry because of the freedom to vary questions on the spot. Frankfort-Nachmias and Nachmias (2008) held that in a non-directive interview the respondent exhibit freedom in describing and defining events and situations in their own words, while the interviewer could probe deeper into the areas of interest. This method is employed in a qualitative study where questions are open-ended to elicit respondents’ interest. A focused interview uses structured questions but allows the respondents degrees of freedom as in non-directive interview.

Personal interview brings the interviewer and the respondent together in a face-to-face interaction and creates a bonding of friendliness between them. Frankfort-Nachmias and Nachmias (2008) argued that personal interview empowers the interviewer to control the process, determine the questions to ask, who to provide the answer, and to decide the location of the interview. They posited that personal interview is more flexible than questionnaire. It provides the opportunity for the interviewer to clarify doubts by probing deeper into the area of interest.

Personal interview however, suffers some limitations which may impede its general acceptability. According to Frankfort-Nachmias and Nachmias (2008) personal interview is more expensive to administer among widely dispersed respondents, is likely to introduce bias, and lacks anonymity necessary to ensure respondents’ freedom. Personal interview however, enables the investigator to learn more from participants’ views than questionnaire.

Telephone Interview

Telephone interview is a substitute to face to face interview. It enjoys real-time high response rate and can reach large respondents within a short time. Frankfort-Nachmias and Nachmias (2008) stated that telephone interview increases the quality of data, and easy to monitor from a central location but that it is less informative as questions are minimized to few important ones. The chance of connectivity breaks off is largely associated with the use of telephone for interview purpose.

The advent of internet offers yet another method of survey data collection through the email. The email is the cheapest method of data collection, with the widest reach and ensures anonymity of respondents but it suffers from most of the limitations of mail questionnaire particularly with regard to low response rate.

Empirical Review of Data Collection Methods

The methods of data collection discussed by Frankfort-Nachmias and Nachmias (2008) are exhaustive and thorough. They were able to analyze each method, stating their peculiar merits and demerits. They also aligned and related the various methods to research designs, showing the peculiar advantage of choosing one over the other. Frankfort-Nachmias and Nachmias (2008) demonstrated knowledge of ethical consideration of using human subjects, when they stated that the main limitation of silent-observer method is that it infringes on the rights of the participants to be informed and secure their willing consents. Frankfort-Nachmias and Nachmias (2008) were not clear with their points when they held that observation is indispensable as data collection method when studying participants who cannot express themselves introspectively or unable to verbalize. The emerging methods of data collection such as internet (e.g. survey monkey) are user-friendly to some vulnerable participants.

Apart from highlighting the ethical consideration bothering on covert observation on the one hand and behavior modification associated with overt observation, Frankfort-Nachmias and Nachmias (2008) did not discuss other limitations of observations such as structural barriers, inaccessibility due to war, hostile cultural differences, and social prejudices which makes observation sometimes impractical. Frankfort-Nachmias and Nachmias (2008) contradicted themselves when they stated that a peculiar advantage of silent observation is that, the investigator does not interrupt the process or the functioning of the people being observed but admitted that the presence of the observer may indeed unintentionally and uncontrollably introduce changes in the events or bias in the process. Frankfort-Nachmias and Nachmias (2008) however, made a valid point about the inability of observation to survey large sample size and may not be well tailored towards a quantitative study where a large sample size is crucial for generalizing findings to a larger population. A restricted sampling of what is observable may not be the true representative of the population and therefore the findings may not be a true reflection of the phenomenon.

Philosophical Assumptions

Philosophical assumptions are the world-views researchers bring into their inquiry. They are the underlying ideas which influence the practice of research (Creswell 2009). The research plan or method chosen by the inquirer is largely driven by the philosophical assumption. According to Creswell (2009) investigators need to identify and clearly express their philosophical ideas when preparing their research plan in other to decide on the appropriate research design. Creswell (2009) posited that philosophical assumption is distinct from strategies and methods of inquiry, but the three intersect in shaping the research proposal.

A research design therefore is comprised of the research strategies, the method of inquiry, and the philosophical assumptions which framed the choice of methods and strategies. Creswell (2009) argued that adequate discussion of the design framework and the interconnections of philosophies, strategies and methods will inform readers why inquirers choose either qualitative, quantitative or mixed methods. A proposal should include sections which sufficiently address the philosophical world-view, the basic ingredients, and the effects on the choice of approach. Creswell (2009) defined world-view as “a basic set of beliefs that guide actions” (p.6). In research parlance, world-view is expressed differently by diverse scholars as epistemologies and ontology, paradigm, or broad research methodologies. Creswell (2009) posited that world-view is the researcher’s orientation of the world and type of research paradigms. According to Creswell (2009) faculties’ and advisers’ beliefs and. disciplines, and student’s exposure and past experiences shape their research world-views.

Research methods refer to the methods of data collections, questions to be addressed, methods of data analysis, interpretation, method of communicating findings and validation. Philosophical world-view predicts the selection of inquiry and questions to be asked, data collection strategies and analysis. The selected strategies of inquiring equally predict the research methods. Creswell (2009) argued that selected research strategies are the qualitative strategies, quantitative strategies and the mixed method strategies. According to Creswell (2009) qualitative strategies comprise of ethnography, case studies, phenomenology, grounded theory and narrative essays approaches. The quantitative strategies may be experimental and quasi-experimental designs while the mixed method strategies include sequential, concurrent, or transformative mixed methods, Creswell (2009) discussed different world-views as post-positivism, constructivism, advocacy/ participatory, and pragmatism.

Post-positivism

Post positivist philosophy emanates from the traditional notion that phenomenon is the outcome of causes and effects. According to Creswell (2009) post-positivism originated from the challenges of the traditional view about the absolute truth in seeking answers to social phenomena. It represents the view that there is no absolute claim to knowledge when it comes to the study of human behaviors. Post-positivist researchers are interested in studying and analyzing causes which influence outcomes through experimental trials. Creswell (2009) stated that post-positivist world-view is also referred to as the scientific method, or empirical science. Post-positivist often adopts the quantitative research design in a reductionist approach by reducing bulk of ideas into small and discrete variables, for testing the hypothesis and answering the research questions.

The post-positivist believes in the existence of laws and theories. They held that laws and theories underline world experiences and are therefore, advanced to be tested to provide an understanding of the problem of study. Post-positivism rests on some key assumptions. These assumptions are the conditions to be met in the choice of research design and provide a guide to the investigators. According to Creswell (2009) a post-positivist philosopher believes that there is no absolute truth about any given phenomenon. This assumption underscores quantitative researchers’ conclusion about the tested hypothesis as research evidence cannot be wholly truthful and infallible. The inquirer cannot prove but at best only fail to reject a hypothesis.

Another assumption is that hence claims are not absolute truth, it should be refined or abandon for emerging valid claims. This is what a quantitative researcher does by advancing the theory as scientific evidence, collect data and analyze same. Researchers express relationships between variables as research questions or hypotheses and statistically assess the validity of the theory or conceptual framework. Creswell (2009) argued that standards of validity and reliability are important qualities of a quantitative research. Objectivity is an essential part of a quantitative inquiry and is a standard guard against bias.

The Social Constructive World-View

Social constructivism underscores qualitative research methods. Often referred to as the naturalistic approach by Lincoln and Guba, (1985) and the social construction of realist by Berger and Luckman as cited in Creswell (2009). It assumes that individuals are in search of understanding their environment. The inquiry is broadened to allow for the study of a wide variety and complexities of views rather than narrowed study of structured variables. The naturalistic inquiry therefore, works through active involvement of participants by asking broad, open-ended and general questions.

Active engagement and interaction with participants characterizes social constructivism. According to Creswell (2009) the process enables the negotiation of subjective meanings. The naturalistic inquiry seeks to understand the meaning of participants about the world by inductively developing theory from emerging patterns. All qualitative inquirers are influenced by social constructivism.

Advocacy and Participatory World-View

Advocacy worldview is predicated on the view that research should address issues of politics, empowerment, inequalities, oppression, dominion, suppression and alienation. According to Creswell (2009) advocacy research is the mouth piece of the oppressed and is aimed at advancing their causes for the improvement of their lives. Creswell (2009) aligned advocacy world view with qualitative research but posited that sometimes quantitative researchers build their research on its principles.

The evolution of this world-view began with the advocatory inquirers feeling dissatisfied that the constructivists have not presented sufficient and robust agenda for the emancipation of the marginalized. The focus of this worldview therefore is on the needs of the marginalized and disenfranchised groups in the society. The philosophical worldview often integrates theoretical perspectives as lens for the construction of a holistic picture of issues being studied. Creswell (2009) posited that inquirers collaborate with participants as active partners in the course of the study. The advocacy aims at creating a paradigm shift in the social spheres leading to positive social change in the lives of the oppressed and vulnerable groups.

Pragmatism

Pragmatism is a world view which focuses on understanding the research problem and adopting all available approaches to produce solutions. A pragmatic inquirer is not restricted by methodological designs. According to Creswell (2009) the concern of a pragmatist is the application of what works to solve problems rather than the method or approach of the investigation. Creswell (2009) held that the world view is a product of burdens for actions and situational consequences instead of focusing on methods. He argued that there are many variations of the pragmatic orientations but all focus on employment of pluralistic approaches to gain knowledge and providing solution to problems. Pragmatic researcher draws from the assumptions of both qualitative and quantitative research in a mixed method design. The pragmatic researcher adopts the mixed methods design, enjoying the synergies of different world-views, different assumptions, and different strategies of data collections and analysis.

Research Approaches

Scientific inquiry is distinct from all other traditional methods of inquiry because it exhibits the highest form of objectivity, reliability, replication, and ability to generalize findings. Frankfort-Nachmias and Nachmias (2008) argued that scientific methodology provides rules for communication, reasoning, and for inter subjectivity because of its appeal to logic and experience. Claims to knowledge by scientists are open to verification by other researchers. Such approach to verification could be either to challenge or confirm earlier assertions, thereby building a body of knowledge in the discipline.

Research method refers to the approach the inquirer employs to conduct empirical investigation. It encompasses the methods of data collection, methods of analysis, and the research design, which flows coherently from epistemological, ontological, and philosophical worldviews. Osuala (1993) argued that research is the process of arriving at dependable and objective solutions to problems through systematic data collection, analysis, and interpretation.

Research aims at discovering the relationship between phenomena. Creswell (2009) held that research design incorporates plans and steps taken to finding answers to research questions. It involves strategies, data collection methods, analysis, and interpretation. Creswell (2009) posited that the worldview assumption informs the decision of choice of design. According to Creswell (2009) worldview is a set of the belief which undergirds the choice of action and methods of investigation. Creswell (2009) identified four worldviews that shape the choice of research methods and design. They are post positivism, constructivism, advocacy, and pragmatism. Research strategies are the specific design or models within the qualitative, quantitative, or mixed methods of research. They are equally called the approaches to inquiry (Creswell 2009). Research methods refer to the forms of data collection, analysis, and interpretation employed by researchers.

Creswell (2009) posited that research method involves different forms of data collection, analysis, and interpretation of research findings, while Plonsky and Gass (2011) noted a positive relationship between the study quality, outcomes, and methodologies. Research method defines the path to follow during the investigation of a given phenomenon. It guides the researcher in the type of question to ask and strategies to adopt in data collection, sampling strategies, and data analysis.

Researchers face the task of choosing the methodologies that best address the problem of study. Crandell, Voils, Chang, and Sandelowski (2011) argued that a major challenge to conducting a scientific research is how to synthesize findings from diverse methodologies. The need to align the method with the problem of study presupposes that researchers must define their choice of methodologies from appropriate philosophical standpoints. According to Creswell (2009), the philosophical worldview of the researcher determines the research method to adopt. Creswell argued that the research questions drive the choice of methodology. The research questions conversely are informed by the philosophical assumptions that underpin the investigation.

The findings of a scientific research must address a problem. Osuala (1993) argued that research is the process of finding solutions to problems through planned and systematic data collection, analysis, and interpretation. Research is aimed at gaining an understanding of a given phenomenon or the discovery of relationships between phenomena.

Quantitative Methods

A quantitative research is an approach situate on post-positivist philosophical world-view. According to Creswell (2009) a quantitative researcher advances theory by evaluating the relationship between variables. The glaring feature of quantitative method is the deductive approach to testing stated theories. The collection and analysis of relevant data is to validate the assumed relationship between the variables. The result of the procedure is either a rejection of the theory, formulating of new one, or the reinforcement of same. A quantitative design employs extensive statistics in the analysis of data to arrive at empirical findings. Creswell (2009) argued that quantitative research is reported in a structured format beginning from the introduction, review of literature and theory, methods, to discussion of results.

The peculiar attributes of a quantitative design are that the researcher exercises control for alternative explanations, generalize findings, randomize participants and minimize chances of bias, promotes objectivity, and build strategies for enhancement of validity. Creswell (2009) defined theory as “a set of interrelated constructs, definitions, and propositions that present a systematic view of phenomena, by specifying relations among variables with the purpose of explaining natural phenomena” (p 51). The definition of theory exposed the focus of a quantitative design as an approach to examine the relationship between constructs or variables. Theories are tested to provide explanations for answers to research questions. Variables in a quantitative research are related, hypothesis tested, and predictions are made.

A quantitative research may be framed in either as a survey or experimental designs. According to Creswell (2009) a survey research is outlined in stating the purpose, identifying the population and determining the sample size, deciding the survey instrument, stating the relationship between variables, stating the research questions and/or the hypothesis, define the method of analysis and the interpretation of data. Generalization is imbedded in the researcher’s ability to control extraneous variables and random assignment of participants. A quantitative method may be conducted in a full experimental or quasi experimental design.

A quantitative research approach involves designing the strategies of data collection, sampling strategies, data analysis and basis for the inferential generalization of finding to the entire population. A quantitative research design may take the form of a survey research, experimental or quasi-experimental design. A quantitative inquirer describes variables before relating them to each other. The researcher compares the groups of independent variables to determine its effects on the dependent variable or relate the independent variables to dependent variables to determine the effect size. He/she evaluates existing theories by gathering descriptive data, analyze them through appropriate statistics in other to make a generalization of the findings to the entire population. The validity of a quantitative design is enhanced by ensuring that every sample within the population stands equal chance of being selected and studied. Any sampling strategy therefore, which does not offer the samples within the population, equal chance of being selected amount to bias.

Research Designs

Research design is the road map for the research as it guides the researcher in data collection, analysis and interpretation. Nachmias and Nachmias (2008) argued that research design allows inferences concerning causal relation and to make valid generalization. They held that experimentation is the classic research design because it enables the researcher to compare variables, manipulates, and control them, and then make generalization. It is when the researcher is able to make comparison that he/she can demonstrate that the variables are related but for him to rule out any other factor as rival explanation he/she must be able to control for such rival variables. The experimental researcher manipulates the process by exercising some form of control over the randomization of the variable and determines the time order in which they are introduced Nachmias and Nachmias (2008).

Experimental Design

Experimental design enables the researcher to study the population by selecting samples, randomized them to groups for treatment, and generalizing the result to the entire population. Creswell (2009) posited that in experimental design, the researcher randomly assigns participants to two groups; administer treatment to one of the groups while isolating the other for comparison and determination of the effect of the treatment. Campbell and Stanley (1966) defined experimental design as that portion of research which enables the inquirer to manipulate variables and observe their effects on other variables. Creswell (2009) held that experimental design is where the researcher assesses the impact of the treatment by administering such treatment to one group while withholding it from another to determine the scores of the two groups on an outcome. According to Creswell (2009) the aim of experimental design is to assess the impact of an intervention on the outcome while controlling for all other factors which might have impacted on such an outcome.

The limitation of experimental design is its incongruence with human social attributes. In experimental design variables are clearly identified and positioned for analysis of their relationships. Independent variables are termed as the predictor variables while the dependent variables are the outcome variables.

To address the issue of bias, the experimenter randomly assigns the participants to groups with each sample standing equal chance of being selected, in other to ensure valid inference and generalization. The experimenter controls for contending variables which may account for the outcome, or intervene in the relationship between the dependent and independent variables. Other variables to be accounted for are the mediating or moderating variables.

Experimental design is that quantitative research approach which enables the inquirer to exercise control and manipulation of variables and random assignment of same to study and experimental groups. The purpose of random assignment is to allow every sample in the population stand equal chance of being selected for study thereby ensuring valid generalization of findings to the entire population. According to Campbell and Stanley (1963) in experimental research variables are manipulated and their effects observed (P.I). Experimental design is most appropriate where the researcher can exercise complete control over variables. It is appropriate when the experiment control enables the scheduling of experimental stimuli Campbell and Stanley (1963).

Assumptions

The assumption of experimental design is predicated on causal and effect relationship between variables under study, and that variables can be assigned between experimental and control groups. Proponent of experimental design, according to Campbell and Stanley (1963), assume that scientific method must be applied to teaching to record progress in the technology of teaching. Campbell and Stanley (1963) argued experimentation sharpens the relevance of testing, probing, and selection process of the traditional approach of investigation. Another assumption of experimentation is that samples selected bear true representation to the population and that by studying the samples findings there from could be generalized to the larger set.

Strength and Weakness of the Design

Experimental design enables the researcher to generalize with degree of confidence and carries with it scientific objectivity and empiricism. Experimentation lowers the degree of type 1 and type 2 errors during scientific investigation. Internal validity according to Campbell and Stanley (1963) is the basic premise for the interpretation of experiment while external validity asses the ability of generalizing experimental findings. A threat to internal validity therefore, impedes the ability to interpret findings while a threat to external validity cripples ability for valid generalization. Campbell and Stanley (1963) stated that internal validity is the ‘sine qua non’ of experimental investigation while strong inference is anchored on external validity. Campbell and Stanley (1963) identified eight threats to internal validity and four threats to external validity.

Quasi-Experimental Design

In a true experimental design, generalization is made valid by the random selection of samples from the population of interest. This makes it possible for selected samples to be true representatives of the entire population. Creswell (2009) however held that in many instances, randomization is not possible as such the experiment will resort to a design called quasi-experiments. According to Creswell (2009) when groups emerge naturally or are not amenable to manipulation, true experiment is not possible. He held that human social attributes are not easily manipulated as such a quasi-experimental design is the best option for quantitative researchers.

A quasi-experimental approach takes the form of non-random assignment of treatment and control groups. Creswell (2009) argued that quasi-experimental designs could take the form of non-equivalent control group designs, simple group interrupted time-series design, or control group interrupted time-series design. While in a true experiment the researcher randomly assigns the participants to treatment groups, in quasi-experimental design randomization is not done, but the investigator utilizes the experimental and control groups. All quasi-experimental design differs from true experiment because of the absence of randomization.

The Cross-Sectional Design

Cross-sectional design is a popular design, employed by social scientists as an alternative to true experimental design. It is mostly aligned with the survey method of data collection. A cross-sectional design is based on data analysis technique of cross-tabulation and bivariate percentage analysis. It is an attempt at addressing the limitation of experimental design by using statistical analysis to approximate the operations inherent in true experiment. Cross–sectional design improves “the design to approximate the post-test only control group”. (Frankfort Nachmias and Nachmias, 2008: 118). The design allows investigators to conduct studies using probability samples thereby increasing the external validity. It provides researchers alternative designs where random assignment of participants to either a control or an experimental group is constrained by ethical consideration. As with all other non-experimental designs, cross-sectional design is weak on internal validity hence I cannot manipulate the independent variable. I will, however employ the logistic multiple regression statistics to increase the power of my analysis.

Qualitative Method

A qualitative design is an exploratory investigation into phenomenon, a method that allows the researcher to explore the subject of investigation with more objectively. The methodologies involve open–ended questionnaire and rigorous data collection and analytical procedures. According to Creswell (2007) a qualitative design is an inductive approach that enables the researcher to learn from participants, while developing theory from bottom–up. The qualitative approach uses the Philosophical views of the constructivists and may take any of five approaches of: Narrative, case study, grounded theory, ethnography or phenomenology. Qualitative research is more focused on verbal data than measurement and more non-numerical data than quantitative research.

According to Experiment-Resources.com (2009) the goal of qualitative research is to help the researcher gather complete and details about the research topic. The data collection strategies are more researcher-involving, as the researcher gather data personally rather than using some form of protocols. Experiment-Resources.com (2009) argued that qualitative design uses different strategies for data gathering such as individual in-depth interviews, structured and non-structured interviews, narratives, documentary analysis, observations and archival retrieval s. Data are presented in words, videos, or image formats, but the approach is more subjective than quantitative research.

According to Schilling (2009) qualitative analysis aims at facilitating the comparison of items to enable valid conclusion to be drawn, but argued that the display of qualitative results often do not provide sufficient answers to problems. Schilling (2009) posited that a more credible and systematic approach to analyzing qualitative data was imperative. The argument by schilling (2009) justified the introduction of ‘concept map’ which is held to be helpful in giving overview of the major themes. The steps to interpreting the ‘concept map’ according to Schilling (2009) involve: Analyzing pattern, clustering objects, making contrasts, and searching for the unvoiced opinions. Schilling (2009) introduced descriptive data analysis to complement the qualitative result and aid valid generalization.

Creswell (2009) posited that qualitative design allows for innovation and creativity, gives room for the researcher to work more within researcher designed framework. He held that qualitative design is often centered on addressing cultural, and societal issues that relates to marginalized group, and focuses on creating a better society.

Qualitative research is more focused on verbal data than measurement and more non-numerical data than quantitative research. According to Experiment-Resources.com (2009) the goal of qualitative research is to help the researcher gather complete and details about the research topic. The data collection strategies are more researcher-involving, as the researcher gather data personally rather than using some form of protocols. Experiment-Resources.com (2009) argued that qualitative design uses different strategies for data gathering such as individual in-depth interviews, structured and non-structured interviews, narratives, documentary analysis, observations and archival retrieval s. Data are presented in words, videos, or image formats, but the approach is more subjective than quantitative research.

Strategies of qualitative research approach

Qualitative research is an exploratory research design which adopts the inductive approach to generate meanings and understanding of a given phenomenon in open-ended strategies. Creswell (2009) posited that qualitative research is situated within a natural setting with unique characteristics of inductiveness, interpretivenes, holistic account, multiple sources of data collection, enabled participants’ meanings, emerging designs, working through theoretical lens and researchers as data collection instruments. Creswell (2009) advocated the five qualitative approaches, arguing that they are popular among social and health sciences methods of inquiry. Nachmias and Nachmias (2008) stated that qualitative researchers seek to understand persons and institutional behaviors through personal contacts as the method of data collection is more closely associated with research individuals through participants’ observation. Creswell (2007) offered the grounds for the adoption of qualitative research would be appropriate for the exploring of a given phenomenon, a need to understudy a group afresh, need for complex and detailed understanding of a phenomenon, need to allow people to freely share their stories, the need to minimize power relations and the need for the findings/ report to be participants inductive.

Grounded theory

The approach involves building new theory through data collection and analysis. Creswell (2009) defined grounded theory as a qualitative strategy in which researchers obtain the views of participants and develops theory from them. Creswell (2007) posited that grounded theory is most needful where a theory does not exist to explain the phenomenon or where it does it is structurally inadequate. Grounded theory approach enables the researcher move beyond description to new sets of theory that offers better explanations of the phenomenon. Creswell (2007) argued that the approach is anchored on the assumption that theory ought to emerge from the analysis of the data collected, studied and analyzed. He identified the systematic and the constructivists strategies of the grounded theory, each defining distinct research procedures. Creswell (2007) presented six stages of data analysis under grounded theory approach to be: open-coding, axial coding, selective coding, development of a conditional matrix and the development of a theory. The need for the researchers to set aside theoretical ideas in other to pave ways for the emergence of theory and the difficulty of determining when categories are fully saturated are the limitations of grounded theory according to Creswell (2007).

Ethnography

Ethnographic approach is most appropriate when the unit of analysis is larger than 20. Creswell (2007) advocated for its use when there is a need to study culturally shared behavioral patterns, beliefs and language of a group of individuals. Ethnographic design according to Creswell (20070 focused on the entire group as it describes and interprets the share values, behaviors and languages. He identified two dominant classes of ethnography to be critical and realist ethnographies. The procedures involve determining its alignment to the research problem its appropriateness, identification and selection of the cultural group, select the type of the ethnographic design to adopt, gather information from participants through lived interactive views of the researcher and the participants. The major limitations of ethnography, Creswell (2007) contended has to do with the need for researchers to have sufficient knowledge of the cultural anthropology of the group, the extended time needed in the field, and the sensitivities of the cultural factors to the participants. Some cultural issues are very personal and very explosive in nature and portend great risk to the inquirer.

Case study

Case study is appropriate for the study of issues through one or multiple cases, adopting detailed data collection and in-depth analysis. Creswell (2007) stated three types of case studies as: simple, multiple and intrinsic cases. The need to identify the specific case to be studied is the difficult task before the researcher. Analysis of case study data may take the forms of holistic or embedded analysis. Creswell (2007) argued that while ethnography seeks to understand how culture works, case study studies issues or problems through case analysis.

Narrative

Narrative focuses on individual stories, arranging them sequentially into meaningful classes using a variety of practical analysis from lived experiences and told stories. Creswell (2007) informed that narrative may come from biographies, autobiographies, told as oral history or life history (memos, journals or diaries).

Phenomenology

A phenomenology is an approach that describes meanings of several individuals lived experiences. It focuses on describing the common experiences of participants expressed as a phenomenon (e.g. grief, discrimination, poverty etc). Creswell (2007) identified the empirical transcendental or psychological phenomenology and hermeneutical phenomenology. According to Creswell (2007) inquirer using phenomenology must state the broad philosophical assumptions and make careful choice of participants to be studied. I personally hold the view that the choice of participants here, as in almost all qualitative approaches is subjective. The researcher is to set aside personal experiences so as not to prejudice his investigation with personal bias.

Mixed Methods

Mixed method is an approach to inquiry that combines or associates both quantitative and qualitative forms (Creswell, 2009). According to Tashakkori and Teddlie, (1998), the basis of the mixed-method approach is the pragmatists’ views that the question addressed by the researcher should indicate the research method(s) used, not the worldview of the researcher or the belief that one method is superior to another. Mixed methods research affords researchers the synergy from the combined strengths of both qualitative and quantitative methods of inquiry. Creswell (2009) held that mixed method blends both quantitative and qualitative assumptions. Researchers use both types of data to provide the best understanding of a research problem.

Researchers have the freedom to choose methods, techniques, and procedures in designing strategies of inquiry. The researcher draws on both quantitative and qualitative approaches and adopts any of the six mixed method approaches.

Sequential Explanatory

Sequential explanatory strategy is used by researchers with strong quantitative inclinations. The strategy is characterized by the collection and analysis of quantitative data followed by the collection and analysis of qualitative data.

Sequential Exploratory

Sequential exploratory researchers, on the other hand, collect and analyze qualitative data before doing the same for quantitative data. They use the quantitative data and results to interpret the qualitative findings. Creswell (2009) argued that this strategy is best suited for testing elements of emerging theory and is used to generalize qualitative findings to different samples.

Sequential Transformative

Sequential transformative strategy is a two phased project with theoretical lenses overlapping the sequential procedure. According to Creswell (2009), the theoretical lenses shape and direct the research question, provide sensitivity for collecting data, and lead to action. Sequential transformative strategy requires that key decisions be made about what findings, in the first phase, will be the focus of the second phase. Further, the absence of sufficient literature on how to use the transformative vision guides the method. Concurrent triangulation strategy requires great effort and expertise to adequately study a phenomenon with two approach methods.

Concurrent Triangulation

A researcher using concurrent triangulation strategy collects both qualitative and quantitative data concurrently and compares them for convergence, differences, or some combination.

Concurrent Embedded

A researcher using concurrent embedded strategy collects data simultaneously from both the primary and secondary sources to address variety of questions.

Concurrent Transformative

A researcher using concurrent transformative strategy uses specific theoretical perspectives as well as the concurrent collection of both qualitative and quantitative data. The researcher could adopt any strategy by drawing on the quantitative and qualitative methods of data analysis, open and closed ended questions, and both emerging and predetermined approaches. A researcher adopts a sequential mixed method to analyze and test data using a build-up strategy of confirming results by adopting either a quantitative followed by qualitative data collection and analysis, as in a sequential explanatory strategy, or the collection and analysis of qualitative data followed by a quantitative approach, as in a sequential exploratory strategy (Creswell, 2009).

The challenges of using mixed methods strategies of inquiry stem mainly from the cost in time and resources. Sequential explanatory strategy is easy to implement, describe, and report, but its main drawback is the length of time involved in data collection. Sequential exploratory strategy suffers from the researcher’s additional responsibility of making key decisions about which findings from the initial qualitative phase will be the focus of the subsequent quantitative phase. Concurrent embedded strategy has the limitation of needing data to be transformed so as to be integrated within the analysis phase of the research. Discrepancies may occur from comparing the two data sets. The concurrent transformative strategy is based on ideologies such as critical theory, advocacy, participatory research, or a conceptual or theoretical formwork. The strategy shares the challenges of triangulation as in embedded approaches (Creswell, 2009).

Conclusion

I have analyzed and evaluated methods and designs for descriptive research and experimental research by Osuala; Data collection methods by Frankfort-Nachmias and Nachmias; philosophical assumptions and quantitative research methods by Creswell, and discovered common trends of constructive and detailed scholarly dialogue by the scholars. According to Osuala (1993) experimental randomization enables generalization of findings to the larger population. This is because it affords every unit in the population equal chance of being selected. Osuala’s (1993) position is in tandem with Creswell (2009) who argued that experimental design is a scientific research method of investigating the relationship between variables. Creswell (2009) addressed the advantages and limitations of experimental design and aligned it to quantitative methods and post-positivist world-views, but Osuala (1993) did not relate experimentation to any research methodologies.

The discussion of experimental design by Creswell (2009) embraced true experiment and quasi-experimental designs but Osuala (1993) was not exhaustive in his analysis of experimentation. Osuala (1993), Frankfort-Nachmias and Nachmias (2008), and Creswell (2009) concurred on the functions of philosophical assumption in driving the choice of research design, methods, and strategies, but Frankfort-Nachmias and Nachmias (2008) elaborated more on data collection method while Creswell (2009) dwelt more on the approaches to empirical research. Osuala (1993) held that observation may take the form of indirect analysis of what was observed by others, but Creswell (2009) disagreed insisting that such approach does not constitute observation but a documentary analysis of secondary data.