Quantitative and Qualitative Paper- Angela driving-FORMATTED TAYLOR AND FRANCIS STYLE No 1

42
A.D. Akorsu and F. Enu-Kwesi A.D. Akorsu and F. Enu-Kwesi Philosophical foundations of social science research versus technical convenience: Methodological weaknesses in students’ researches Angela Dziedzom Akorsu 1 Institute for Development Studies, University of Cape Coast, Cape Coast, Ghana & Francis Enu-Kwesi 2 Institute for Development Studies, University of Cape Coast, Cape Coast, Ghana 1 Institute for Development Studies, University of Cape Coast, Cape Coast,Ghana. Email: [email protected] , [email protected] , [email protected] . 2 Institute for Development Studies, University of Cape Coast, Cape Coast,Ghana. Email: : [email protected] , [email protected] , [email protected]

Transcript of Quantitative and Qualitative Paper- Angela driving-FORMATTED TAYLOR AND FRANCIS STYLE No 1

A.D. Akorsu and F. Enu-Kwesi A.D. Akorsu and F. Enu-Kwesi

Philosophical foundations of social science research versus technical convenience: Methodological weaknessesin students’ researches

Angela Dziedzom Akorsu1 Institute for Development Studies,University of Cape Coast,Cape Coast, Ghana

&

Francis Enu-Kwesi2

Institute for Development Studies,University of Cape Coast,Cape Coast, Ghana

1 Institute for Development Studies, University of Cape Coast, Cape Coast,Ghana. Email: [email protected], [email protected],[email protected].

2 Institute for Development Studies, University of Cape Coast, Cape Coast,Ghana. Email: : [email protected], [email protected], [email protected]

AbstractSocial science research is replete with studies that use or

combine quantitative and qualitative research strategies.

Yet, quantitative and qualitative research strategies are

driven by different epistemological and ontological rhetoric

and also give rise to differences in the kind of

intellectual knowledge that can be produced. By means of a

review of the justification often presented by graduate

students for using particular methods, this paper provides

several examples which illustrate that using or combining

quantitative and qualitative strategies for purposes of

triangulation is often flawed. Without dismissing

compatibility claims between quantitative and qualitative

research strategies or the integrationist approach, this

paper argues that the choice of any approach should be

justified and informed, not only by technical

considerations, but by philosophical and analytical

considerations – seeking answers to different types of

research objectives and thereby serving complementary

purposes.

1

Key words: epistemology, ontology, quantitative, qualitative, mixed method, methodology, research, social

sciences

Introduction

The quantitative-qualitative debates have moved from one of

contradiction to one of compatibility. Some writers (Lincoln &

Guba, 1985; Noblitt & Hare, 1988; Rosenberg, 1988) have

questioned the wisdom of integrating both approaches due to their

different purposes. According to these authors, the quantitative

2

paradigm in social science research is underlain by the study of

social variables that lead to the prediction of social

phenomenon. The researcher’s responsibility is thus to quantify

these social variables and phenomena in order to be able to

predict future events or outcomes. Qualitative paradigm, on the

other hand, beckons greater understanding rather than prediction

of social phenomena. Hence, the doubters of integration argue

that the two paradigms differ so much that any integration or

reconciliation is likely to undermine the epistemological

foundations of each (Rosenberg, 1988).

Others, however, while appreciating the philosophical

debates, do take pragmatic positions (Cook & Reichardt, 1979;

Steckler, 1989; Steckler, McLeroy, Goodman, Bird & McCormick,

1992) and suggest that nothing should stop any social scientists

from subscribing to the philosophy of one paradigm while

employing the methods of the other paradigm. The integrationist

proponents argue that the complexity of social interventions such

as those that occur in water, sanitation, health, poverty and

education programmes require the usage of multiple or mixed

methodologies in order to produce effective evaluation of these

interventions.

Beneath the arguments and counter arguments for the adoption

of single methodology approaches, it is worth taking into account

the strengths and weaknesses of each approach and the fact that

single methodology approaches often fail to explore all the

3

components of social phenomena. This failure can and sometimes

renders well-intentioned social interventions less effective than

desired. Jones (1997) admits that the crucial justification for a

mixed methodology design is that both qualitative and

quantitative methodologies have strengths and weaknesses. It is

thus incumbent upon the researcher to blend both methods so that

the final research outcome can highlight the important

contributions of each one, where qualitative data can support and

give meaning to quantitative findings.

The objective of mixed methods is not necessarily to

reconcile qualitative and quantitative approaches, but rather to

note that they are not antithetical and that the practical

exigencies of the problem or issue being addressed may require

the application of both methodologies (Pinto, 2010). What can be

gleaned from Pinto’s view is that as social scientists we aim to

understand the complexity of human behaviour and as researchers,

our task is to describe and explain this complexity. However,

when there are limitations in our individual methodological

repertoires, it is beneficial to acknowledge that and adopt other

methods or mix methods through triangulation (Neuman, 2011). In

doing so, the dimensions and scope of the problem at hand can be

broadened and elucidated to generate a holistic perspective of

human behaviour and experience, which in turn can lead to

effective solutions for social problems.

4

Similar views have been expressed by other researchers.

Bowen (2005) explained how her dissertation committee encouraged

her to adopt the best of both approaches in order to be able to

reflect the interactions and experiences of individuals and

communities in respect of her chosen research problem. According

to Bowen, both approaches have distinct and complementary

strengths. In her contribution to the debate, Hastings (2010)

refers to methodological triangulation or the use of multiple

methods to study a single problem. In this respect, qualitative

and quantitative methods may be employed simultaneously or

sequentially. According to Hastings, this can occur either as

within-methods triangulation where multiple quantitative or

qualitative approaches are employed, or as between-methods

triangulation where both qualitative and quantitative approaches

are used. Post positivists argue that triangulation enables

researchers to minimize biases that are associated with single

methodological approaches, and constructivists assert that

triangulation is beneficial since it allows the discussion of

multiple perspectives of the problem and necessitates the

researcher to consider multiple realities.

In this paper we argue that the underlying philosophies of

qualitative, quantitative and mixed methods research strategies

cannot be ignored if the knowledge produced is to be of value.

Consequently, the usage of specific methods must be justified

with acceptable assumptions and with appropriate and adequate

analytical procedures. The rest of the paper is organized into

5

five sections. Section two dwells copiously on the differences

between intensive (qualitative) and extensive (quantitative)

research in relation to epistemological assumptions, ontological

assumptions and intellectual goals while section three presents

an argument for and against hybrid approach. Section four

describes in brief, the methodology adopted, while section five

comprises evidence from graduate students’ researches regarding

the appropriateness of the selected methods. The last section

contains the conclusions.

The quantitative-qualitative debates

A research strategy is a model or framework that guides an

entire research agenda. According to Inkoom (1999), it refers to

the logical sequence that connects the empirical data to the

study’s initial questions and ultimately to its conclusions.

Blaikie (2000) adds that a research strategy, which can also be

referred to as logic of enquiry, provides a starting point and

steps by which questions can be answered. According to Harwell

(2011), research strategy is important because it communicates

information about the key features of the study and these can

differ for qualitative, quantitative and mixed methods. What

this means is that, the kind of research strategy used will have

implications for the kinds of epistemological and ontological

assumptions that can be made, as well as implications for the

kinds of methods that can be used and more importantly,

6

implications for the kinds of intellectual knowledge that can be

produced.

Clarke (2005) cites Creswell’s (2003) reference to the two

main categories of research paradigms that are often employed in

social science research, as qualitative and quantitative

research strategies, which are also sometimes referred to as

intensive and extensive research strategies respectively.

Regarding these, there have been considerable debates as to which

one is superior and a more desirable choice for social inquiry.

Some researchers view quantitative strategies as superior. In the

words of Lord Kelvin, “when you cannot measure it, when you

cannot express it in numbers, your knowledge is of a meagre and

unsatisfactory kind” (Sayer, 1992, p. 175). This perception has

received a lot of reactions, including Jacob Viner’s that “when

you can measure it, when you can express it in numbers, your

knowledge is still meagre and unsatisfactory” (Sayer, 1992,

p.175). Although intensive research design emerged later, with

time, it has come to be an established alternative to extensive

research strategy (Denzin & Lincoln, 2005).

According to Harwell (2010), while the distinction between

intensive and extensive research strategies superficially can be

just a question of scale or ‘depth versus breadth’, in actual

fact, they ask different sorts of questions, use different

techniques and methods and define their objectives and boundaries

differently. This superficiality is reflected in Sayer’s 1992)

description of the choice between intensive and extensive

7

research strategies as a “dilemma” and it is indeed a dilemma

when the underlying philosophies are not put into proper

perspective. This confirms Bryman’s (1988, p. 3) statement that

“increasingly, the terms ‘quantitative research’ and ‘qualitative

research’ came to signify much more than ways of data collection;

they came to denote divergent assumptions about the nature and

purpose of research in the social sciences”.

Extensive research strategy is an earlier strategy and often

shown as scientific since it has characteristics of the natural

sciences. The tendency has been to perceive it as more scientific

and therefore more accurate than intensive research. According to

Clarke (2005), the extensive or quantitative approach, is often

called traditional, positivist, experimental or empiricist and

was advanced by authorities such as Comte, Mill, Durkheim, Newton

and Locke. It is rooted in positivist assumptions, having a

logical structure in which problems are derived from theories. A

common example is the survey design which seeks to measure and

evaluate perceptions of sampled individuals that are thought to

be representative of the larger population to which the results

can be generalized (Kraska, 2010). This strategy lends itself to

the use of statistical packages for easy manipulation of data

into aggregation, categorization, correlation, regression and

hypothesis testing.

Intensive or qualitative research strategy on the other

hand, has come to be perceived as a more socially oriented

approach. This strategy, according to Tashakkori and Teddlie

8

(2003), has introduced a lot more innovative methods like

critical incident technique, diary studies, discourse analysis

interviews, life histories, grounded theory, document analysis,

case studies, cognitive mapping, ethnography, narratives,

metaphors, participant observation and repertory grids for

answering social questions. It takes a constructivist,

naturalistic, interpretive, postpositivist, and postmodernist

perspective as advanced by eminent writers like Dithey, Kant,

Foucalt, Miles and Huberman (Clarke, 2005).

The intensive strategy employs a number of methodological

approaches, and is driven by diverse epistemological assumptions.

It involves exploration of social relations and describes reality

as experienced or constructed by the respondents (Sarantakos,

1993, 2005). Due to its constructionist and interpretivist

theoretical foundations, the use of intensive strategy requires

considerable painstaking effort (Sarantakos, 2005), because

prior categories are not adopted and statistical tools that

manipulate data are uncommon.

The underlying assumptions of extensive and intensive

researches in turn lead to differences that are outside the realm

of philosophy. According to Sale, Lohfeld and Brazil (2002, p.

45), “the two paradigms have given rise to different journals,

different sources of funding, different expertise and different

methods. They state that there are even differences in scientific

language used”. For example, “validity” to the extensive

9

researcher means that a research finding is exactly the same as

the existing reality out there, whereas to the intensive

researcher, “validity” is used for a research account that is

agreeable.

Epistemological differences

To a very large extent, extensive research strategy is

rooted in a positivist orientation which assumes that social

facts are objective and exist outside the influence of the

individual while intensive research strategy is driven by a

number of epistemological claims - phenomenological,

hermeneutical, interpretive and naturalist claims - with the

overall belief that reality is socially constructed by

individuals as well as by groups (society). One tenet of positive

analysis which is retained in all of the various forms of

positive philosophy of science is the emphasis on empirical

evidence or some statements, which must at least in principle, be

testable (Caldwell, 1992; Creswell, 2009; Sarantakos, 2005). This

feature of positivism is called empiricism and it implies that,

certain characteristics of social objects such as feelings and

language which are different from the objects of natural sciences

and cannot be observed are not considered as objects for study

since scientific methods cannot be applied to them. The

difficulty with this view however, lies in its over-ambitious

search for existing empirical evidence, which makes it

unrealistic.

10

Throughout the history of science, empirical evidence has

not been able to conclusively prove or falsify scientific

statements because the demands for verification are too extreme

and scientific observation is unavoidably “theory-laden”.

Empirical evidence has at best, led to only an approximation of

truth. The intensive paradigm on the other hand acknowledges the

fact that certain attributes of social objects are unobservable

but can still be studied if the researcher is interpretive and

reflexive (Creswell, 2003, 2009).

Another positivist characteristic of extensive research

strategy is that methods of the natural sciences are applied to

the social sciences so that any discipline which does not lend

itself to the methods of natural science is not considered as

science. This principle is referred to as the principle of

methodological monism or methodological naturalism (Giedymin, 1975; Von

Wright, 1971). For example, the extensive research strategy is

deductive and almost always starts with a theory, from which

hypothesis is stated. Data are then collected and analyzed to

either support or reject the stated hypothesis. This way of

conducting research, according to Kuhn (1984), is the way of

conducting research in natural and physical sciences and seems to

suggest that experiments and/or empirical studies are about

confirmation rather than about discovery.

However, in arguing for the intensive approach, Sayer (1992,

p. 123) is of the opinion that “sciences and their methods vary

according to the nature of their objects”. Thus, applying the

11

methods of natural sciences to the study of social objects

misses the point because social objects are unique and have

certain characteristics that cannot be studied by the methods of

natural sciences. A good method should therefore reflect a good

understanding of the nature of the objects under study, and this

can be accomplished through intensive approaches.

The treatment of values in positivism is also evident in

extensive research approaches. Harwell (2011) discusses that

integral to the extensive approach is the expectation that

researchers will set aside their experiences, perceptions and

biases to ensure objectivity. This view on one hand treats values

as potentially harmful to objectivity in research. To emphasize

objectivity and ensure the validity of the knowledge produced,

there must be ways of restricting the inference of the

researcher’s values in the research process. A good extensive

research “ aims to achieve a form of scientific objectivity in

which data remain uncoloured by the theoretical baggage,

especially the values and interests, that researchers bring to

collecting and interpreting data” ( Howe, 1992, p. 239). A good

intensive research, on the other hand, must be value-laden, and

the values and experiences of the researcher are actually

necessary for the study of social objects. This implies the

acceptance of multiple truths that are socially constructed

(Harwell, 2011).

The role of inductivism and deductivism as regards theory in

extensive research strategies is also a reflection of the

12

positivist orientation. On one hand, theories are developed by a

constellation of verified facts and on the other hand, theories

are the drivers for the search for empirical evidence. This seems

to imply that the research process is circular rather than

linear. Bryman (1988, p. 21) however, posits that in reality,

there is a “lack of a clearly ordered sequence of steps in

quantitative research ... quantitative research is invariably

much more messy”.

The alternative to induction and deduction strategies in

extensive research is the ‘Abduction’ strategy in intensive

research (Blaikie, 2000). According to Blaikie, abduction is a

process of generating social scientific accounts from the

accounts of social actors and is associated with interpretive

approaches. Blaikie’s description of abduction here appears to

have the characteristics of grounded theory which was described

as “the most influential paradigm for qualitative research in the

social sciences” (Denzin, 1997, p. 18). It was originally

presented as systematic, inductive and informed by positivism

with its associated emphasis on objectivity. Charmaz (2003)

however, states that its positivist premises have come under

serious attack, and argues that it has more constructivist

tendencies than positivist tendencies. It is within the

constructivist stance of grounded theories that similarities with

abduction can be placed.

13

Ontological differences

The differences in epistemological assumptions between

extensive and intensive research strategies have implications for

a number of ontological assumptions. The role of the researcher

is one area of ontological difference between intensive and

extensive research strategies. In quantitative research the

researcher must be detached from the research process so as to

avoid biases (Creswell, 2009; Harwell, 2011; Sarantakos, 2005).

For instance in the administration of questionnaires in a survey,

researchers often employ field assistants and are often not

directly involved in the data collection. Intensive research by

contrast requires much more contact since the contact enables the

researcher to have reliable basis for interpretation and

understanding of the objects. The researcher must actually be

immersed in the object of study. This means that the researchers’

values, experiences and perception of the object are useful in

explaining the objects of study (Harwell, 2011; Wahyuni, 2012).

Subjectivity is a feature of intensive research as opposed to the

emphasis on objectivity in extensive research strategy.

In the area of research design, it can be seen that extensive

research is highly structured, with processes that are

systematic. As a strategy, extensive research concentrates on

seeking empirical evidence by emphasizing the validity and

reliability of methods of data collection and analysis and

rejecting subjectivity of the researcher. In view of its

interest in generalization, larger samples are preferred to

14

smaller samples, and a detailed and careful sampling is required

as a way of ensuring representativeness. By contrast, intensive

research is associated with flexibility and lack of structure.

This is also evident in the kinds of methods adopted such as

critical incident technique, diary studies, discourse analysis

interviews, hermeneutics, life histories, grounded theory,

document analysis, case studies, cognitive mapping, ethnography,

narratives, metaphors, participant observation and repertory

grids. Since generalization is not the ultimate goal, small

samples or even single cases or incidents or stories are adopted

because they are technically realistic.

The nature of data is also different between extensive and

intensive research. The data from extensive research is regarded

as hard, rigorous and accurate, and comes in the form of numbers,

tables, graphs and statistical symbols and jargons, similar to

data from the natural sciences. The rigor and accuracy are often

used to suggest superiority even though such data are often

meaningless to many people – literate and illiterate. On the

other hand, by virtue of what intensive research seeks to

achieve, its data are usually rich with detail and insightful in

meaning.

In extensive research, there is considerable focus on

individuals as the unit of inquiry as occurs in survey designs

where questionnaires and interview schedules are administered to

individual respondents and the results aggregated. This is what

Bryman (1988) calls individualism as opposed to holism in

15

intensive research, where the unit of inquiry is often the whole

– school, organization, slum, marriages etc. While the issue of

the individualism in extensive research appears the case, in the

long run, the individual cases and characteristics become

submerged in a pool of aggregated analysis. In a similar vein,

while intensive research +

seeks to study a phenomenon as a whole, in reality, abstractions

are made. It is almost impossible to explain a social object as a

whole without carefully abstracting from its constituents first.

Though both quantitative and qualitative research discuss

issues of causality, some differences are worth noting.

Quantitative researches are preoccupied with establishing the

causal relationship between variables – cause and effect

analysis. The use of “dependent” and “independent” variables in

extensive researches is indicative of this fact and is also an

importation of positivist mentality of treating social objects as

natural science objects in closed systems. Causality in intensive

research is not as simple as that since objects exist in open

systems and do not produce automatic regularities. According to

Sayer (1992), “causality concerns not only a relationship between

discrete events (cause and effect), but the causal powers and

liabilities of objects or relations ... if the nature of an

object changes then its causal powers will change too”. The

implication of this is that, the flexibility inherent in

intensive research is required in studying causality in social

objects.

16

Differences in intellectual goals

Taylor and Bogdan (1984) summarize the differences in

intellectual goals between extensive and intensive research as

follows: quantitative research seeks to explain the causes of

changes in social facts, primarily through representative

sampling, objective measurement and quantitative analysis, with

generalization and prediction as possible goals. According to

Bryman (1988), this preoccupation with establishing generality

and predictability can probably be attributed to the tendency of

the extensive researcher to mimic the methods and style of the

natural sciences. In reality, generalization is hardly accurate

in the social sciences and Sayer (1992) calls the goals of

generalization and predictability a mistaken one.

Sayer (1992) distinguishes between predictive explanations

and explanatory predictions and posits that due to the stable

nature of objects in closed systems, explanatory predictions can

be derived. For instance once there is night, there is bound to

be day break. The opposite is the case in open systems which lend

themselves for non-predictive explanations (causation accounts)

such as the theory of evolution and there is no apparent

justification for assuming that such non-predictive explanations

are incomplete explanations. Ultimately, the aim of intensive

research is to explain social objects in a way that provide an

in-depth understanding of the reality as is constantly

constructed by social objects.

17

Qualitative research is more concerned with understanding

the social phenomenon from the actors’ perspectives through

participation in the lives of those actors. As opposed to

generalization, intensive research has contextualism and

understanding (Verstehen) as goals. Objects are studied and

understood within specific contexts. Thus characteristically, an

intensive research is often based on single case analysis or

small sample sizes with no particular emphasis on

representativeness but the uniqueness and the essential features

of each case is upheld as significant for understanding the

social phenomenon. .

Closely related to the issues of generalization and

contextualization is the issue of replication, which is given

much more prominence in extensive research with the focus on

cross-checking an earlier study for any excesses in the

researcher’s biases and other interruptions. Intensive research

on the other hand is difficult to replicate given its emphasis on

the unique context as well as the influence of the

‘idiosyncrasies’ of the researcher. A plausible intellectual goal

therefore may be corroboration rather than replication. Thus, in

view of the emphasis on representation, generalization and

replication, the ultimate intellectual goal of extensive research

is to produce what Bryman (1988) and Babbie (2007) call

‘nomothetic’ reasoning where general laws are established and

regarded as applicable even to different times and places. The

emphasis on context in intensive research is also said to result

18

in ‘idiographic’ reasoning or locating the findings within

specific time and place (Babbie, 2007).

Another intellectual goal worth discussing in extensive

research is quantification. This works to eliminate the study of

a host of important variables such as socialization processes

through which we acquire our gender identities just because they

cannot be quantified. Thus, quantification and the use of

mathematical language in extensive research leads only to

calculating, deducing and deriving conclusions based on

assumptions rather than seeking structural and causal

explanations of a phenomenon (Sayer, 1992). Alternatively,

description in intensive research, coupled with examination of

theory (Sayer, 1992) is important in highlighting the richness

and depth of the phenomenon under study, since social objects

tend to be concept-dependent and have multiple intrinsic

meanings. While description can be criticised for its lack of

intellectual rigour, Rist (1984, p. 161) asserts that “it’s at

once disarmingly simple and incredibly complex”. In reality

however, qualitative research often goes beyond mere description

to provide analyses of the objects and environment of study.

Regarding the intellectual goal of making theoretical

claims, extensive and intensive researches differ as well. While

in extensive research, the position of theory is strongly

entrenched very early in the study, so that any empirical

evidence obtained is either to accept or reject the theoretical

claims (Saunders, Lewis & Thornhill, 2009), in intensive

19

research, theory is not given such prominence at an early stage.

In intensive research, theory and empirical evidence are

intertwined (Neuman, 2011). This accords with Bryman’s (1988, p.

81) description that “the delineation of theoretical ideas is

usually viewed as a phrase that occurs during, or at the end of

the field work, rather than being a precursor to it”. The

manifestation of this is in the use of grounded theory where

theories are derived from the field work process and also, what

has been called analytic induction in intensive researches.

Combining extensive and intensive research strategies

Despite the conflicting epistemological and technical

differences between the extensive and intensive research

strategies (Neuman, 2011; Saunders et al., 2009; Wahyuni, 2012),

there have been several arguments for combining the two

approaches. First, it has been argued that a preoccupation with

the extensive-intensive debate is futile as it has no promise of

a solution in the near future and also because ‘epistemological

purity’ does not get research done (Miles & Huberman, 1984). This

view is supported by Howe (1988) who posited that truth is what

works and so what works out well is what researchers should do.

Secondly, it has been said that in view of the fact that the two

approaches are both legitimate ways of understanding the world,

they can be combined (Haase & Myers, 1988). Thirdly, combining

20

the two approaches is useful in addressing issues of complexity

in the phenomenon under study (Sale et al., 2002).

Migiro and Magangi (2011) argued that when used in

combination in one study, quantitative and qualitative methods

complement each other and allow for a more complete analysis of

the research problem. According to them, the approaches are

complementary rather than competitive and the use of a particular

method or the decision to use both methods in a single study must

be based on the nature of the actual research problem. This

position is in symphony with the views of Lincoln and Guba (2005)

as expressed by Christ (2007) that studies should be flexible,

contrary to Yin’s (2006) argument that researchers must use

preconceived procedures including overarching research questions

that cover both the quantitative and qualitative aspects of the

study.

Regarding the reasons for using mixed methods, the most

frequently cited reason is that of triangulation and cross-

validation. Triangulation has been described by Denzin (1970) as

an approach in which multiple observers, theoretical

perspectives, sources of data and methodologies are combined. The

logic of triangulation is to ensure validity and the emphasis on

validity in this way is a positivist claim. Thus, it is important

to note what worldview is guiding the overall study so as to

ensure epistemic consistency (Creswell, 2003; Denzin & Lincoln,

2005). Validity as used here is not a major aim in intensive

21

research. Another reason for using a mixed method is for

facilitation - using one as a precursor for the other. In this

regard Hussein (2009) argues that triangulation is more precise

as it aims to reveal complementarity, convergence and dissonance

among the findings.

According to Morgan (1998), the most frequently used style

is one that starts with a qualitative pilot study, followed by

quantitative research as the main study. Rather than argue over

which approach is best Condelli and Wrigley (2004) suggest a

movement away from what they call “false dichotomy” and

gravitation toward mixed methods. According to them, by using a

rigorous design the quantitative or extensive approach can tell

us what works, while the qualitative or intensive approach can

tell us how it works. Our position is that in applying extensive,

intensive or mixed methods the researcher needs to provide

justifications and apply analytical procedures that will make the

end product acceptable. The methods used in collecting the data

must fit the selected approach, and it is appropriate for the

researcher to explain any tilts where the mixed method is

applied. This makes it easier for other researchers to evaluate

or utilize the information thus contributed in a fitting context.

Methodology

Research works undertaken by students at the Institute for

Development Studies of the University of Cape Coast provided

22

material for examining the usage of the various approaches and

how such selected approaches were justified. The total number was

not available, so all theses written since the inception of

graduate programmes in 1996 and which had been displayed on the

library shelves were used. The displayed theses were examined for

the approaches and justifications. The examination also focused

on whether the methods used in collecting and analyzing data were

appropriate for the chosen approaches.

For convenience, the selected theses were grouped into three

tables based on five-year intervals, starting from 1996. Table 1

contained information on theses from 1996 to 2000, while Tables 2

and 3 contained information on theses submitted in 2001-2005 and

2006 to 2009 respectively. The displayed theses spanned the

period 1996 to 2009. According to the attending librarians, at

the time of this study, the School of Graduate Studies and

Research had not released theses submitted in 2010, 2011 and

2012.

Empirical evidence from students’ researches

For purposes of anonymity and confidentiality, the names of

the student researchers have been omitted from the tables, and

key words have been used to represent the titles. The dates have

been maintained to provide a chronology of the theses and to help

23

readers appraise whether the quality of the methodologies and the

analytical rigour has improved, stayed the same or deteriorated

over time. The numbering of the thesis has been continued from

table to table in order to allow readers to easily know the

number of theses examined. The contents of the theses were

examined with respect to research design and the specific study

design, and how the choice was justified and at what stage in the

process the justification occurred. Regarding mixed method,

attention was paid to the particular chapter in the thesis that

it was mentioned. This was to determine whether the students were

following procedures or were just acting conveniently. The

examination also focused on analytical tools and their

appropriateness.

Table 1. Students’ theses submitted to the Institute for Development Studies, University of Cape Coast for the period 1996-2000No Date Topic,

KeywordsResearch Design Design Data

Justification

Where mix mentioned

Analysis

1 1996 Rural banks,impact

None stated Not Stated , not mentioned

Frequencies, cross-tabs, chi-square

2 1996 Application of new technologies to farmers

Not Stated Not Stated , not mentioned

Frequencies and percentages

3 1996 Rural water mgt,training

Evaluative,descriptive

Not Stated ,

Percentages

24

not mentioned

4 1996 Early childhood care and development-programmes

Participatory Approach

Not Stated , not mentioned

Frequencies, cross-tabs

5 1997 SSI, environmentalconcerns

Case study Not Stated , not mentioned

Frequencies, cross-tabs, chi-square

6 1999 Development of technologicalcapabilities

Interpretive paradigm, casestudy, grounded theory

Indirectly within methodologychapter

Interpretative, frequencies, means, variance

7

1999 Safety training programme evaluation

Experimental design

Not Stated , not mentioned

Descriptive and inferential

8 2000 Population,poverty, envt degradation

Mixed Not Stated , not mentioned

Frequencies, percentages

9 2000 Patent system,industrial propertyand economic dev.

Qualitative Not Stated , not mentioned

Frequencies, percentages

Source: Institute for Development Studies’ Library, 2012

The evidence from Table 1 show that frequencies andpercentages, interspersed with chi- square analysis from cross-

25

tabulations were the predominant analytical tools that were usedby students in the beginning years of the programme. Theseanalytical tools were chosen, even when the topic and the designsuggested different methods or procedures. While some of thestudent researchers specified mixed methods yet chose frequenciesand percentages to present the data, others clearly stated thatthey used qualitative, participatory or case study approaches butwent ahead and used frequencies and percentages. These analyticaland presentation choices are very inconsistent and undermine theessence of the chosen study designs.

Table 2. Students’ theses submitted to the Institute for Development Studies, University of Cape Coast for the period 2001-2005

No Date Topic, Keywords

Research Design Design Data

Justification

Where mix mentioned

Analysis

10 2001 Women’s production and the envt.

Not stated Not Stated , not mentioned

Frequencies, percentages, cross-tabs

11 2001 SAP and crop production

Model specification

Not Stated , not mentioned

Multiple regression

12 2001 Community perception,education

exploratory Not Stated , not mentioned

Frequencies, Percentages, chi-square

13 2001 Rural water systems,

Descriptive, cross

Not Stated ,

Problem tree

26

management sectional not mentioned

analysis,objectivetree analysis,SWOT

14 2001 SSE’s, linkage patterns

Sample survey Not Stated , not mentioned

Frequencies, percentages, means

15 2002 Community participation

Mixed Justified In data processing and analysis

Frequencies, percentages

16 2002 Technologicalcapacity building, Biotechnology

Mixed Not stated

Data analysis

Descriptive, interpretative, frequencies, percentages

17 2002 Sanitation,hygiene

Triangulation,lab analysis

Not Stated , not mentioned

Descriptive, laboratory analysis

18 2002 Innovation dissemination

Mixed Not Stated , not mentioned

Frequencies, percentages

19 2002 Women’s circumstances

Desc., cross-sect, evaluative.

Not Stated , not mentioned

Frequencies, content analysis

20 2002 NGOs, street children

Not stated Not Stated , not mentioned

Descriptive, comparative statistic

27

s21 2002 Communication

, project planning and implementation

Exploratory Not Stated , not mentioned

Frequencies, means

22 2002 Attitudes to girl-child education

Exploratory, Mixed

Not Stated , not mentioned

Frequencies, chi-square

23 2003 Local Development Financing, DA’s, UE,UW

Mixed (case study)

Not stated

Data analysis

Regression, t-distribution

24 2003 Sustainable Rural Livelihoods, local partnership

Mixed(PRA/case study)

Not stated

Method of study

Sustainable livelihood analytical framework, descriptive, frequencies

Table 2 continued

25

2003 Credit programme for women

Case study Not Stated ,not mentioned

Frequencies, percentages, chi-square

26

2003 Community Participation, Poverty

No design stated

Not Stated ,not mentioned

Frequencies, chi-

28

reduction prog.

square, correlation

27

2003 SSEs, promotion

Comparative design

Not Stated ,not mentioned

Frequencies, chi-square, Anova

28

2003 Labour needs

Qualitative Not Stated ,not mentioned

Interpretative

29

2004 Recruitmentand Retention, health workers

Exploratory/descriptive

Not Stated ,not mentioned

Frequencies, percentages, chi-square

30

2005 Price fluctuations, maize

Mixed Not stated Data analysis

Problem tree analysis

31

2005 NGOs and women’s groups

Descriptive survey

JUSTIFICATION; Qualitative datareflects researchdesign

Frequencis, percentages

32

2005 HR and capacity building needs

Cross-sectionaldescriptive

Not Stated ,not mentioned

Frequencis, percentages, means

Source: Institute for Development Studies’ Library, 2012

As can be seen or inferred from Table 2, similarmethodological errors and inconsistencies were prevalent inresearches that were undertaken by students more than five yearsafter the implementation of the programme. Apart from a fewstudents who ventured into different types of analysis, most of

29

the researchers concentrated on using frequencies and percentageswith associated chi-square analysis.

Based upon the evidence in Table 3, it is quite obvious thatthe passage of time did nothing to raise the knowledge of thestudent researches regarding the choice of approach. The analysisremained weak and inconsistent with the chosen approaches. Wheremixed methods were specified, very little was seen of any realqualitative analytical procedures. For reliability andacceptability of the knowledge created some of the topics shouldhave been researched qualitatively, using constant comparative,phenomenology or ethnography as the basis of analysis, yet thestudents resorted to the usual frequencies and percentages.

Table 3. Students’ theses submitted to the Institute for Development Studies, University of Cape Coast for the period 2006-2009

No Date Topic, Keywords

Research Design Design Data

Justification

Where mix mentioned

Analysis

33 2006 Child nutrition and school performance

Desc, cross-sect.Quantitative

Not stated,not mentioned

Frequencies, percentages, chi-square

34 2006 TraditionalAuthoritiesin local governance

Not stated Not stated,not mentioned

Frequencies, percentages

35 2006 Community participation, sustainableDevelopment

Not stated Not stated,not mentioned

Interpretative, frequencies

30

36 2007 Hygiene Behaviour change

Not stated Not mentioned

Frequencies, percentages

37 2007 Enterpreneurial skill development

Case Study Not stated,not mentioned

Frequencies, percentages, chi-square

38 2007 Women, SSI,AEED

Quantitative Not stated,no mentioned

Frequencies, percentages

39 2007 Adolescent sexual and Reproductive health awareness

Mixed Stated in Research design p59. Justification based on Sarantakos,1998

Frequencies, percentages, chi-square

40 2007 Fertility behavior

Cross sectional Not stated,not mentioned

Frequencies, cross-tabs, Pearson correlation

41 2007 Indigenous institutions and development

Mixed Justified in study design

Discourse, contentanalysis,frequencies, percentages

42 2008 World Vision in Rural Develoment

Quantitative(Descriptive survey)

Not stated Data analysis

Frequencies, percentages

43 2008 Traditional ‘more ‘ so Comparati

31

authorities, Role in rural development

qualitative’ analysis began on the field’

ve analysis,frequencies, percentages

44 2008 Asset categories,poverty, land

Mixed (case study)

Not stated,not mentioned

Chi-square

45 2008 Surface mining and the environment

Mixed (cross-sectional descriptive)

Not stated,not mentioned

Chi-square

46 2009 Training Programmes,Headteachers

Not stated Not stated,not mentioned

Frequencies, means

47 2009 Training needs assessment

Case Study Not stated,Data analysis

Frequencies, percentages

48 2009 Migration and status

Descriptive/explanatory

Not stated,not mentioned

Frequencies, percentages, averages

Source: Institute for Development Studies’ Library, 2012

What we noticed from the review of the various theses

available showed an overall weakness in the methodologies adopted

for the studies. Furthermore in the cases where both qualitative

and quantitative approaches were used the researchers simply

reduced each one to its elementary form and as Steckler et al.

(1992) recommended, the researchers should have increased the32

level of sophistication for each method in order to make the

knowledge created more reliable. For those students who wrote

that they used qualitative studies, notwithstanding what they had

written, none actually used any of the typical data collection

procedures, which according to Tashakkori and Teddlie (2003),

should include a lot more innovative methods like critical

incident technique, diary studies, life histories, grounded

theory, document analysis, narratives, metaphors, ethnographic

field notes and participant observation.

The researchers who used the quantitative approach relied

mostly on data presentation techniques like frequency tables and

charts, percentages and proportions and occasionally ventured

into measures of central tendency. Even those who reported

measures of central tendency could not attach any sophistication

as to which measure could appropriately be referred to as the

average. The accompanying measures of dispersion were also left

un-discussed. Usually, a quantitative approach, using survey

instruments like questionnaires and interview schedules, should

include inferential analysis like estimation and hypothesis

testing (Steckler et al., 1992; Sarantakos, 2005; Babbie, 2007).

These analyses should be done at both the composite level and a

disaggregate level, but all were absent.

Four PhD researchers used mixed methods but the

justification for their use were not provided and mention of

quantitative or qualitative was usually in the section on data

33

analysis, instead of research design. Eight M Phil researchers

stated that they used mixed methods but only four offered

justifications and these could be found, either in the research

design section or in the data analysis section. Mostly no

justifications were offered and neither was the mix method even

mentioned. The analytical techniques were just as simple as those

used by the researchers who adopted single methodology

approaches, quite contrary to what Steckler et al. (1992) advised

that mix methods warrant more advanced analysis. In fairness to

the researchers, many adduced reasons of inadequate time and

funds for their inability to undertake more advanced analysis.

Despite those reasons, it is plausible that their inability to

apply more advanced analytical techniques is due more to

unfamiliarity and personal methodological weaknesses.

Conclusions

The available literature on quantitative and qualitative

research strategies suggests first of all that the two strategies

are both legitimate strategies towards social inquiry. Thus, the

traditional view that quantitative research is more scientific

and therefore more desired has not been sustained. The

qualitative strategy can be an innovative and viable alternative

to quantitative research strategy. They both have their

respective places in social science research, depending of course

on the phenomenon under study. Secondly, while acknowledging the

differences between the two strategies, many modern writers on34

the subject feel strongly that whatever the differences between

the two perspectives may be, a synthesis between the two

strategies is desirable - mixing the two strategies in a single

study for various reasons but with triangulation as the most

prominent reason.

However, as have been portrayed in this paper, the

differences between intensive and extensive research strategies

are not trivial at all. They range from objectivity versus

subjectivity, fixed categories versus emergent categories,

outsider versus insider’s perspectives and a static reality

versus fluid reality. Clearly, these differences go beyond

technical issues to include epistemological and ontological

assumptions as well as differences in intellectual goals. The

choice of one over the other as well as any combination of them

should therefore be informed, not only by technical

considerations, but by philosophical considerations as well. They

ask different types of research questions – each serves a certain

unique purpose. For this simple reason, extensive and intensive

research strategies should not just be mixed in a single study

for triangulation or cross-validation purposes only. Any attempt

to use a hybrid of the two approaches should be appropriately

justified and appropriately applied, taking into account the

analytical expectations for each.

It is apparent that many students are not aware of the

appropriate analytical tools for either quantitative studies or

qualitative studies, let alone the mixed methods approaches. It

35

is important for the role that each method plays to be made clear

in the research process and this should be backed by proper

implementation. This has implications for the way that research

methods courses should be taught and how courses that deal with

data analysis ought to be handled. The implications also extend

to theses’ supervision and assessment since these are integral

part of students’ knowledge creation processes. It is incumbent

upon all those who are involved in the processes to insist upon

clarity of the approach and consistency in the paradigm.

References

Babbie, E. (2007). The practice of social research (11th ed.). Belmont,CA: Wadsworth Cengage Learning.

Blaikie, N. (2000). Designing social research: The logic of anticipation.Cambridge: Polity Publications.

Bowen, G. A. (2005). Preparing a Qualitative Research-Based Dissertation:

Lessons Learned. The Qualitative Report, 10 (2), 208-222.

Bryman, A. (1988). Quantity and quality in social research. London:Routledge.

Caldwell, B. (1982). Beyond positivism: Economic methodology in the twentiethcentury. London: Unwin Hyman Limited.

Cassell, C., & Symon, G. (Eds.) (1994). Qualitative methods in organizational research. New York: Sage Publications.

36

Charmaz, K. (2003). Grounded theories: Objectivist and constructivist methods. London: Sage Publications. Christ, T. W. (2007). A recursive approach to mixed methods research in a longitudinal study of postsecondary education disability support services. Journal of Mixed Methods Research, 1, 226-241. DOI: 10.1177/1558689807301101.

Clarke, R. J. (2005). Research models and methodologies. UoW HDR Seminar Series. Faculty of Commerce, Spring Session 2005.

Condelli, L., & Wrigley, H. S. (2004, March). Real world research: Combining qualitative and quantitative research for adult ESL. Paper presented at the National Research and Development Centre. Second International Conference for Adult Literacy and Numeracy.

Loughborough, England. March 25-27, 2004.

Cook, T. T., & Reichardt, C. S. (1979). Qualitative and quantitative methods in evaluation research. Beverly Hills, CA: Sage Publications.

Creswell, J. W. (2003). Research design: Qualitative, quantitative and mixed methods approaches (2nd ed.). London: Sage Publications.

Creswell, J. W. (2009). Research design: Qualitative, quantitative and mixed methods approaches (3rd ed.). London: Sage Publications.

Denzin, N.K (1970). The research act in sociology. London: Butterworth.

Denzin, N. K., & Lincoln, Y. S. (Eds.). (2005). The Sage handbook ofqualitative research (3rd ed.). Thousand Oaks, CA: Sage Publications.

37

Giedymin, J. (1975). Antipositivism in contemporary philosophy ofsocial science and humanities. British Journal of the Philosophy of Science,26, (4) 275-301.

Harwell, M. R. (2011). Research design in qualitative/quantitative/mixedmethods. Chapter 10. www.sagepub.com/upm-data/41165_10.pdf

Haase, J. E., & Myers, S.T. (1988). Reconciling paradigmassumptions of qualitative and quantitative research. WesternJournal of Nursing Research, 10, 128-137.

Hastings, S. L. (2010). Triangulation. Encyclopedia of Research Design.Sage Publications Inc.

http://www.sage-ereference.com/researchdesign/Article_n469.html.

Howe, K. R. (1988). Against the quantitative-qualitativeincompatibility thesis or dogmas die hard. Educational Research,17, 10 – 16.

Howe, K. R. (1992). Getting over the quantitative-qualitativedebate. American Journal of Education, 100, 236-257.

Hussein, A. (2009). The use of triangulation in social sciencesresearch: Can qualitative and quantitative methods becombined? Journal of Comparative Social Work, 1, 1-12.

Inkoom, D.K.B. (1999). Management of non-reserve forest in Ghana: A casestudy of Mpohor Wassa East District. SPRING Research Series 24, Facultyof Spatial Planning, University of Dortmund, Germany.

Jones, I. (1997). Mixing qualitative and quantitative methods insports fan research. The Qualitative Report, 3 (4).(http://www.nova.edu/ssss/QR/QR3-4/jones.html)

38

Kraska, M. (2010). Quantitative research. Encyclopedia of ResearchDesign. Sage Publications Inc. http://www.sage-ereference.com/researchdesign/Article_n352.html.

Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. NewburyPark, CA: Sage Publications.

Lincoln, Y. S., & Guba, E. G. (2005). Paradigmatic controversies,contradictions and emerging confluences. In N. K. Denzin, & Y.S. Lincoln (Eds.), The SAGE handbook of qualitative research (pp. 192-214). Thousand Oaks, CA: Sage Publications.

Migiro, S. O., & Magangi, B. A. (2011). Mixed methods: A reviewof literature and the future of the new research paradigm.African Journal of Business Management, 5 (10), 3757-3764.

Miles, M., & Huberman, A. (1984). Drawing valid meaning fromqualitative data: Toward a shared craft. Educational Researcher,13, 20-30.

Morgan, D.L. (1998). Practical strategies for combiningqualitative and quantitative methods: Applications to healthresearch. Qualitative Health Research, 8, 362-376.

Neuman, W. L. (2011). Social research methods: Qualitative and quantitativeapproaches (7th ed.). New York: Pearson.

Noblitt, G. W., & Hare, R. D. (1988). Meta-ethnography: Synthesizingqualitative studies. Newbury Park, CA: Sage.

Pinto, R. M. (2010). Mixed methods design. Encyclopedia of ResearchDesign. Sage Publications Inc. http://www.sage-ereference.com/researchdesign/Article_n245.html

39

Rists, R.C. (1984). On the application of qualitative research topolicy process: An emergent linkage. In L. Barton, & S. Walker (Eds.), Social crisis and educational research (pp. 153-170). London: Croom Helm.

Rosenberg, A. (1988). Philosophy of social science. Boulder, CO: Westview Press Inc.

Sale, J.E.M., Lohfeld, L.H., & Brazil, K. (2002). Revisiting the quantitative-qualitative debate: Implications for mixed-methods research. Journal of Quality and Quantity, 36 (1), 43-53.

Sarantakos, S. (1993) Social research. Sydney: Macmillan Press.

Sarantakos, S. (2005). Social research (3rd ed.). New York, NY: Palgrave Macmillan.

Saunders, M., Lewis, P., & Thornhill, A. (2009). Research methods for business students. London: Pearson Education.

Sayer, A. (1992). Methods in social science: A realist approach. London: Routledge.

Steckler, A. (1989). The use of qualitative evaluation methods totest internal validity: An example in a work site health program. Evaluation and the Health Professions, 12, 115- 133.

Stickler, A., McLeroy, K. R., Goodman, R. M., Bird, S. T., & McCormick, L. (1992). Toward integrating qualitative and quantitative methods: An introduction. Health Education Behavior, 19, 1. DOI: 10.1177/109019819201900101. http://heb.sagepub.com

40

Tashakkori, A., & Teddlie, C. (Eds.).(2003). Handbook of mixed methdos in social and behavioural research. Thousand Oaks, California: Sage Publications.

Von Wright, G. H. (1971) Explanations and understanding. London: Routledge.

Wahyuni, D. (2012). The research design maze: Understanding paradigms, cases, methods and methodologies. JAMAR, 10 (1), 69-80.

Yin, R. (2006, April). Mixed methods research: Are methods genuinely integrated or merely parallel? Paper presented at the annual meetingof the American Educational Research Association, San Francisco.

41