University of Pretoria etd – Wanyama, M N (2006)

Chapter 3: Research Design and Methodology

CHAPTER THREE
RESEARCH DESIGN AND METHODOLOGY
3.1 INTRODUCTION
This chapter details the methodology employed in the study. Among
the areas discussed under research design are descriptive and
diagnostic research approaches, and case study research along with
its qualitative dimension as reflected in this study. Data collection
methods such as the interview method, focus group discussions,
observation methods and content analysis are discussed in detail as to
how they were used and the extent of their effectiveness. Finally, the
scope and limitations of the study are examined.

According Mindoti (1999) research methodology in ethnomusicology

borrows a lot from natural science methodology. However, the natural
sciences mainly emphasize the testing and establishing of the laws of
the universe using experimental methods that can be repeated several
times for empirical verification of the previous results as a way of
ensuring objectivity. For ethnomusicology, the essence is in the
generation of knowledge based on creativity and the interpretation of
behavior in a given context. For example, ethnomusicology
researchers do not derive meaning from mere observation of behavior,
they go ahead and discuss the context in which the behavior has
occurred, that is, the cultural factors influencing the producer in order to
make a meaningful interpretation of the behavior under observation. In
other words, observation makes meaning only when the factors
influencing the producer of the music are considered in order to
achieve objective results. Above all, unlike the natural sciences that
use objects, figures and chemicals in experimenting on the laws of
nature, ethnomusicology, as Mindoti (Ibid) observes, uses human
beings as ‘specimen’ to generate knowledge. This makes the research
methodology in ethnomusicology intricate and difficult to achieve the
type of objectivity, which the typical scientific methodology aims to
achieve.

3-1
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

This study relied on two complementary sources of data: primary and

secondary. In order to situate the study theoretically and generate the
conceptual framework with which to work on the primary sources,
secondary data was consulted and analyzed to provide the initial data
for the study. Secondary evidence included written sources like books,
journal articles, daily newspapers, government reports, articles,
seminar papers, MA and PhD theses etc. These were accessed from
libraries at Kenyatta University, Moi University, University of Nairobi
and the Jomo Kenyatta Memorial Library among others. Online
journals available on Internet were also consulted.

Primary written documents on the topic were gathered from the Kenya
National Archives. These are the first hand records from the
participants and observers that have been passed down to posterity.
Government monographs, ministry of culture reports, native council
reports etc. were consulted for information relating to this research.
Visits were made to the Kenya Broadcasting Corporation Library, the
Institute of African Studies, East African Standard and Nation
Newspapers Library, Music Copyright Association of Kenya etc. in
order to refer to the recordings and the information available.

Another primary source was the carrying out of both participant and
non-participant observations and interviews mainly within Bungoma
district. Purposive sampling technique was used to identify
respondents including circumcisers, initiates, song leaders,
instrumentalists etc. among other key informant cultural consultants.
Focused in-depth individual and group interviews were carried out.

Recordings on audiocassettes and videotapes captured group

performances and individual interviews. Six research assistants were
trained and employed to help in the research process. Participation,
observation and the use of an interview schedule provided the

3-2
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

researcher with the opportunity to have the actual field experience and
raise further questions.

The data collected from documentary sources and fieldwork was

qualitatively analyzed. Here the information on audiotapes and
videotapes was transcribed and interviews summarized and coded to
come up with clear understandable statements and conclusions. The
music collected was classified and analyzed according to the various
stages of the Bukusu circumcision ceremony. This was done by
counterchecking; comparing, contrasting and corroborating the
information collected from various sources together with the theoretical
framework outlined, research questions, hypothesis and objectives.

In order to gather the information required, oral interview schedules,

observations and focus group discussions were utilized. Tape
recorders were used during interviews to get information directly
without much trouble of struggling to write it down; and hence
disrupting the flow and coherency of information from respondents.
This helped instill confidence in the interviewees. While using the oral
interview schedule, the researcher’s probing and prompting gave room
for greater depth of response. Where necessary and appropriate,
English, Lubukusu, and Kiswahili languages were used
interchangeably. Moreover, still photo cameras and videotapes were
used to capture some of the participants (dancers, initiates, and
singers), dance formations and venues of music performance (see the
CD and DVD accompanying this thesis).

3.2 RESEARCH DESIGN

This section describes research design and methodology in terms of
population, sampling, and administration of research instruments, data
collection procedures, and the description of techniques used in data
analysis all of which were utilized in order to answer the research
questions and achieve the set objectives. The grammatical and
speculative theories form the theoretical basis of the study. The

3-3
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

conceptual perspective of the study is based on qualitative research

that includes designs, techniques and measures that do not produce
discrete numerical data. According to Mugenda and Mugenda
(1999:155), more often the data in qualitative research are in the form
of words rather than numbers and these words are grouped into
categories. They further observe that human behaviour is explained
best by using qualitative research. Human phenomena that cannot be
investigated by direct observation such as attitudes and other emotions
are best studied using the qualitative method. In order to achieve the
research objectives, the research design was based on descriptive and
diagnostic research approach. On the role of qualitative research in
information transfer Mugenda and Mugenda further observe:
Finally, emerging issues relating to social, political
and economic development in poor countries have
enhanced the use of qualitative approaches in search
of sustainable solutions to the myriad problems facing
these countries. Some research and evaluation
experts have argued for the qualitative approach
especially in Africa because communities in Africa
have traditionally communicated information by word
of mouth rather than in written form. The older
members of the community are considered to have
wisdom and it is their obligation to pass on this
wisdom to the younger generation. Folklore, for
example, has been an effective framework of
communicating information, especially on issues
related to sex and family life, religious beliefs, taboos,
sickness, social mythology etc. Because of the
tendency of African communities to pass information
orally, there is a strong argument that the most
appropriate research and evaluation approach in
Africa is the qualitative approach because it
emphasizes oral communication and gives
respondents a chance to state their problems the way
they perceive them and participate in seeking
solutions to these problems as well as in effecting
such solutions (1999: 202).

3.2.1 DESCRIPTIVE AND DIAGNOSTIC RESEARCH APPROACHES

In the current research, most of the procedures appertaining to
descriptive and diagnostic research approaches were adhered to. To
begin with, an oral interview schedule was pre-tested by using it as a

3-4
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

guide in interviewing several people who were familiar with Bukusu

circumcision customs. This pilot study was conducted in Moi University,
Eldoret town and Bungoma district prior to the actual circumcision
ceremony period. Consequently, the gaps detected in the schedule
were appropriately filled in order to reflect on the entire scope of the
study in finer details.

3.2.2 CASE STUDY RESEARCH: THE QUALITATIVE DIMENSION

According to Gillham (2004:10), “qualitative methods are essentially
descriptive and inferential in character and, for this reason, are often
seen as ‘soft.’ You may have significant statistical results, but these
have to be described and interpreted: ‘facts’ do not speak for
themselves - someone has to speak for them.’’ This is a strong
argument that justifies the use of qualitative methods to answer some
questions in the current research. In this case, the philosophical base
is that human behavior; thoughts and feelings are partly determined by
their context. The researcher understands the Bukusu people in real
life situations and by studying them in their context and in the way they
operate he managed to situate the place of Bukusu circumcision music
and hence make sense of it in terms of rules, contexts, functions,
meanings, value and significance. This contextual approach led to the
achievement of research objectives with ease. In further support for the
qualitative approach, Gillham observes that ‘[o]bjectivity’ can ignore
data important for adequate understanding (Ibid. 2004:12). Therefore,
on the basis of the foregoing, this study adopted the qualitative
approach in the collection and analysis of data. A researcher-
administered oral interview schedule that was used as a guide in
controlling the direction of interviews was in most cases effectively
used. However, it is worth noting that due the discursive and
descriptive nature of the qualitative approach it was often time
consuming.

3-5
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

3.2.3 RESEARCH STRATEGY

This study employed two research strategies that were inter-twined.
These are ethnography and case study. Ethnography has its origins in
the work of anthropologists studying aspects of a particular group in
depth. It involves the researcher becoming a member of the group
being studied in order to share their experiences and try to understand
why members of the group act in particular ways. In view of the fact
that the focus of this study is on the performance of traditional Bukusu
circumcision music, which is a culture-specific phenomenon,
ethnographic and case study strategies were more applicable.

3.3 DATA COLLECTION METHODS

3.3.1 INTRODUCTION
Case study was the main method within which other methods such as
interviews, focus group discussions, observations and content analysis
were utilized in the context of the study as explained below.

3.3.2 TRAINING RESEARCH ASSISTANTS

Six research assistants (see appendix no. 6) were trained by the
researcher on how to administer an oral interview schedule. Each of
them was given ample time to go through the research proposal before
discussing it in detail with the researcher. The researcher had a
session with all the six research assistants where he illuminated the
perspective of the study as a social research. Research assistants
raised pertinent contentious issues concerning the content and
interpretation of various questions. Technical terms used in the
questionnaire such as form, content, performance, composition etc.
were clarified and exemplified by the researcher in relation to the
context of the study. Furthermore, the researcher discussed with
assistants the research objectives and hypotheses in detail. From the
discussions it became necessary to formulate other guiding questions
in addition to the interview schedule. This was a deliberate attempt to
curb the foreseen digressions by the informants. Lastly, the researcher
enlightened his assistants on the effective use of questioning

3-6
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

techniques, participant observation, and non-participant observation

and on general issues concerning ethics in social research but with a
specific inclination to the perspective of the current study. He also gave
them guidelines on the effective use of still photo cameras, video
cameras and audio tape recorders. After the theoretical and
psychological preparation, the researcher equipped them with the
necessary stationery and set them ready for fieldwork.

3.3.3 INTERVIEW METHOD

In this study interview method was employed when conducting key
informant interviews and in focus group discussions as illustrated
below. The interviews were guided by an oral interview schedule as
discussed below.

3.3.3.1 Oral Interview Schedule

The current research used the unstructured interview schedule that is
sometimes called an interview guide. In this kind of interview, the
interviewers asked questions or made comments intended to lead the
respondent towards giving data to meet the study objectives. Because
of the open nature of unstructured interviews, probing was commonly
used to get deeper information.

3.3.3.2 Key Informant Interview

A key informant interview is used to collect data from persons who are
considered to hold crucial and relevant information. In this case the
targeted respondents were among others, elderly members (elders) of
the society who have experienced various changing phases of the
practice of the Bukusu circumcision over a substantial period of time.
Others were the initiates’ parents, local administration officials such as
headmen, councillors and chiefs (see photo no. 3.1). Apart from
contributing other crucial information on the research topic, the local
government officials also gave out views concerning the current Kenya
government policy stand in regard to traditional Bukusu circumcision
rite. They also articulated issues to do with the relationship between

3-7
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

the spread of HIV/AIDS and the traditional circumcision rite. Also

interviewed were circumcisers who contributed information about the
ceremony and the current emerging changes.

Photo no. 3.1; taken by the researcher: A Local Government

administration official, Caleb Maseti of Bahai Bungoma District
being interviewed.

3-8
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

The interview schedule had open-ended questions focusing on the

research objectives. It was left open to enable the interviewees to
express their perspectives. More often they would come up with a new
perspective not captured in the instrument. The style of the interview
was conversational, and the objectives were achieved through active
engagement by the interviewee and the interviewer. An interview
schedule with open-ended questions differs from the more structured
composition and uniform style of the survey interview (Mason, 2002).
The approach made the interview interactive, situational and had a
generative approach to the acquisition of data. The questions in the
schedule required the respondents to divulge information on culture as
a changing phenomenon in reference to the organization and practice
of the Bukusu circumcision ceremony in which the music under
investigation in this study is part and parcel. All the respondents
targeted for these data collection methods were successfully
interviewed.

3.3.3.3 Focus Group Discussions

According to Litosseliti (2003:15), ‘‘focus group methodology has
developed as a result of broader shift from quantitative to qualitative
research methods. It is important for qualitative research and can
combine flexibility and adaptability with rigor and theoretical
grounding.’’ In this study, focus group discussions were used to gather
views and opinions by giving participants a topic or a series of
questions to discuss. The researcher’s main role in the discussions
was to listen and moderate the discussion by probing and prompting
views from the participants. The researcher used this methodology
successfully by discussing research questions with groups of a few
purposively sampled respondents such as circumcisers, initiates’
parents, Bukusu culture consultants, composers, singers and dancers.

3.3.4 OBSERVATION METHOD

The observation method is the most commonly used method especially
in studies relating to behavioural sciences. Kothari observes that:

3-9
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

In a way we all observe things around us, but this sort

of observation is not scientific observation. Observation
becomes a scientific tool and the method of data
collection for the researcher when it serves a
formulated research purpose, is systematically planned
and is subjected to checks and controls on validity and
reliability (1990:118).

These views are corroborated with Binnet’s (2003:97), when he notes

that ‘‘observation is used as a research technique when data on actual
practices are required. Like questionnaires, observation generates
factual information rather than explanations.’’

As indicated above, the purpose of participant observation is to

develop an insider’s view of the setting and persons under study.
Nevertheless, Patton (1990) advocates the combination of participation
and observation because it gives the researcher the possibility of
understanding the activity as an insider while describing the same for
outsiders. In the current study, the researcher and research assistants
took part in learning and performing the circumcision music. However,
it was at times necessary for the researcher to get detached and have
‘a birds eye view’ of the activities from an angle especially when
recording some crucial information or taking a picture of a particular
dance, movement or sub-ritual. This technique assisted in avoiding
gaps, which would arise from the participatory approach where the
researcher would not be able to notice and capture gestures and
performance patterns going on behind him.

3.3.5 CONTENT ANALYSIS

According to Achola in Mwiria and Wamahiu (1995:46), ‘‘content
analysis is more appropriately defined as a research technique for
systematic analysis, qualitative, quantitative or both, of the manifest or
latent meanings of words, phrases, ideas or postures, objects or
artifacts.’’

3-10
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

This research technique involves a detailed study of documents

relevant to a study. Such documents took a number of forms, including
books, local newspapers, existing databases of information, journals,
relevant theses, audio and videotapes.

3.4 SAMPLING
In most ethnomusicological research, it is neither possible to determine
the number of informants nor songs that give a fair representation of
music in a given culture. Similarly, Merriam (1964:54) argues that
‘‘cutting across the two areas of field and laboratory techniques is the
extremely difficult question of what constitutes an adequate sample of
the music of a community, tribe or larger grouping.’’ He further argues
that ‘‘creativity is a never ending process under whatever culture rules
it is carried out; thus what might be conceived as a total sample one
day, may be lacking the next. What percentage, then, of an infinite
sample constitutes reliability? The answer is that there is no answer.’’

As stated above, creativity in music is an on-going process, which is

infinite, and as such, it is difficult to establish the amount of music,
which is an adequate representative sample. However, in this study,
the researcher used two non-probability sampling techniques namely:
purposive sampling and snowball sampling. Kerlinger (1973:129)
observes that ‘‘non-probability samples are often necessary and
unavoidable. Their weaknesses can to some extend be mitigated by
using knowledge expertise, and care in selecting samples and by
replicating studies with different samples.’’ The current research
integrated purposive and snowball sampling that led to a
complimentary/symbiotic relationship as shown in figure 3.1.

3-11
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

FIGURE NO. 3.1: Researcher’s Illustration of the Interface between

Purposive and Snowball Sampling Techniques.

PURPOSIVE
RESEARCH SNOWBALL
SAMPLING
SAMPLE SAMPLING

3.4.1 PURPOSIVE SAMPLING

Mugenda and Mugenda (1999:50) observe that purposive sampling is
a sampling technique that allows a researcher to use cases that have
the required information with respect to the objectives of his or her
study. Therefore, subjects are hand picked because they are
informative or they posses the required characteristics. In this study,
the sample size of subjects consisted of four initiates, four initiates’
parents, two circumcisers, three song leader/composers, two local
government officers and two key Bukusu cultural informants who were
purposively identified through snowball sampling. These together with
other participants made up a total of 36 subjects that were interviewed
at the end of the study (see appendix 9).

3.4.2 SNOWBALL SAMPLING

According to Mugenda and Mugenda (1999:51), in this method, initial
subjects with the desired characteristics are identified using purposeful
sampling technique. The few identified subjects name others that they
know have the required characteristics until the researcher gets the

3-12
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

number of cases he or she requires. In this study, the researcher used

snowball sampling by identifying some of his Bukusu students in Moi
University (where the researcher teaches) whose brothers; relatives or
neighbours were to be circumcised. This was done as from January to
march, 2004. These students informed the initiates’ parents or
guardians of the intended study and its general objectives. Between
April and June, 2004, the researcher organized meetings with the
initiates’ parents with whom he discussed the issue and further
requested them to identify key informants, who mainly included elders
from the respective localities that were knowledgeable in the study
area. Moreover, the researcher requested the initiate’s parents to
identify local government officials such as headmen, chiefs and
councillors who had crucial and relevant information in regard to this
study.

3.5 DATA VALIDITY PROCEDURES

Perakyla (1997) and Leninger (1994) point out that it is possible to
ensure validity of qualitative data if the researcher follows procedures
that minimize biases. In the present study, several strategies were
used to ensure the accuracy of the findings. Data collected through
different techniques like observation, interviews and focus group
discussions was triangulated to build coherent justification for various
themes. The researcher and assistants compared notes at the end of
every data collection to ensure that what was recorded was not subject
to personal bias. Also, tape recorded data was used in cases where
there was uncertainty or incomprehension of the respondents’ actual
words.

Since the research assistants were known within the study area, most
respondents were willing to provide the necessary information. The
research assistants also followed similar interview guides and the
researcher impressed upon them to keep accurate and thorough field
notes.

3-13
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

The collected Bukusu circumcision music was transcribed by using

staff notation. This notation system has some limitations because it
leaves out a number of inherent aspects of the Bukusu circumcision
music. For instance, it does not have conventions for ululation and
other vocal embellishments. Due to this fact, recorded music on audio
cassette tapes; and still photos were used for the purpose of
representing broader aspects of the music. In addition to the audio-
recorded music, dance, movement, speech, paramusical and
paralinguistic features were captured on videotapes and on digital
videodisks (DVD). It is worth noting that in the performance of Bukusu
circumcision music, in most cases, soloists have the freedom to
extemporize with regard to rhythm and melody while repeating their
phrases. In such cases, only the varied repeats were transcribed.

3.6 DATA PROCESSING AND ANALYSIS

After the fieldwork, data collected from documentary sources and the
field was qualitatively analyzed. Here the information on audiotapes
and videotapes were edited and interviews summarized and coded to
come up with clear understandable statements and conclusions. The
songs collected from the field were classified and analyzed according
to the various stages of the Bukusu circumcision ceremony. Descriptive
data and evidence relating to each research question were classified
into distinctive classes based on their common qualitative
characteristics. Being basically a qualitative research, the results were
discussed in a narrative manner. For purposes of quick reference, each
line in every song is numbered chronologically.

3.6.1 UNITS OF ANALYSIS

Mugenda and Mugenda (1999) observe that the units of analysis, also
called the units of statistical analysis refers to those units that we
initially describe for the purpose of aggregating their characteristics in
order to describe some larger group or abstract phenomenon. In the
current study, the main units of analysis are derived from the form,
content and performance of the Bukusu circumcision music.

3-14
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

Specifically, the main units are: Identification of the music as vocal,

instrumental or vocal with instrumental accompaniment; basic forms,
phraseology, harmonic principles and styles,
performance/presentational form; and lastly, the nature and meaning of
the text. The theoretical procedures/frames, as employed in the study,
are fully illustrated in figure 3.2. In the figure, the theoretical
procedures/frames are based on the grammatical and speculative
theories that form the theoretical framework of the entire study as
discussed in chapter two (section 2.3).

FIGURE NO. 3.2: Researcher’s illustration of theoretical

procedures/frames derived from Nissio Fiagbedzi’s “Philosophy
of theory in ethnomusicological research” in Djedje, J.G. and
Carter, W.G. (eds.). (1989): Volume I. pp. 45-57.

THEORETICAL PROCEDURES /FRAMES

GRAMMATICAL
SPECULATIVE
PROCEDURE
‰ Rules
PROCEDURE
‰ Form ‰ Meanings
‰ Content ‰ Significance
‰ Context ‰ Aesthetics
‰ Functions ‰ Values
‰ Language ‰ Functions
use

3-15
 University of Pretoria etd – Wanyama, M N (2006)
Chapter 3: Research Design and Methodology

3.7 SCOPE AND LIMITATION OF THE STUDY

The main units of analysis in this study were: the dominant solo-and-
response song structure form, visual arts, verbal arts, dance and the
creative-cultural-philosophical aspects underpinning the composition
procedure and performance of the Bukusu circumcision music. Apart
from being notated, the analyzed songs were translated from Lubukusu
(the language spoken by the Bukusu people) to English language. It is
worth noting that this research did not cover the notation of dance.
However, descriptive approach was used to explain the basic dance
formations evident in the performance of Bukusu circumcision music.
Because the nature of the study is more of an indigenous knowledge
product, qualitative research was the most suitable as it was necessary
to explore complexities that were beyond the scope of more controlled
approaches, which are usually based on quantity than quality.

Lastly, it is worth mentioning that the presence of electronic equipment

such as still photo cameras, cassette recorders and video cameras led
to respondents or performers shying away from the interviews or
exaggerating their actions and/or gestures.

3.8 CONCLUSION
In this chapter, the research methodology pertinent to the study is
discussed. In sum, focus is put on research design; descriptive and
diagnostic research approaches, the qualitative dimensions of case
study research and research strategy. The effectiveness of the two
sampling methods used namely: snowball and purposive sampling, is
also discussed together with data validity, processing and analysis
procedures. Moreover, data collection methods such as interview,
focus group discussions, observations and content analysis are
evaluated in terms of how they were effectively used. Lastly, while
discussing the scope and limitations of the study, delimitations of the
study and its impediments are illuminated.

3-16