After completing the introduction and literature review, the next task in the proposal is the methodology. This chapter of the research thesis deals with the description of the methods applied in carrying out the research study. It is organized under the following sections: research design, research site, population, sampling techniques, research instruments, data collection procedures and data analysis.
A research design can be thought of as the structure of research. It is the “glue” that holds all of the elements in a research project together. A design is used to structure the research, to show how all of the major parts of the research project work together to try to address the central research questions. Orodho (2003) defines it as the scheme, outline or plan that is used to generate answers to research problems. A research design can be regarded as an arrangement of conditions for collection and analysis of data in a manner that aims to combine relevance with the research purpose. It is the conceptual structure within which research is conducted. It constitutes the blueprint for the collection, measurement and analysis of data (Kothari, 2003).
Types of Research Designs
It is important to understand the relationship among various designs. This will assist the researcher in making design choices and thinking about the strengths and weaknesses of different designs. The designs are as follows:
The major purpose of descriptive research is description of the state of affairs as it exists. The researcher reports the findings. Kerlinger (1969) points out that descriptive studies are not only restricted to fact findings, but may often result in the formulation of important principles of knowledge and solution to significant problems. They are more than just a collection of data. They involve measurement, classification, analysis, comparison and interpretation of data.
Descriptive survey is a method of collecting information by interviewing or administering a questionnaire to a sample of individuals (Orodho, 2003). It can be used when collecting information about people’s attitudes, opinions, habits or any of the variety of education or social issues (Orodho and Kombo, 2002). For example, teachers in schools can carry out a survey to find out students’ attitudes towards their teaching styles or discipline. When using this design the researcher should ensure the following:
- Construct questions that will solicit the desired information.
- Identify the individuals that will be surveyed.
- Identify the means by which the survey will be conducted.
- Summarize the data in a way that provides the designed descriptive information.
In this design, subjects are randomly assigned to an experimental group which receives the treatment or to a control group which does not receive treatment. Assuming the two groups were initially equivalent, the researcher can compare their performance. In this design cause and effect can be easily determined. If you decide to use this design, you must be certain of the independent and dependent variables and must guard against the influence of extraneous variables.
This design enables the researcher to assess the degree of relationship that exists between two or more variables. It analyzes .the correlation between two or more variables (Orodho, 2003). For example, if you compare the examination performance of a group of university. students who prepare their own meals everyday and those who eat at the cafeteria, you will use a correlation design. Suppose the academic performance of students who prepare their own meals is lower than those who eat at the cafeteria, then you may deduce that preparation of meals by students has an impact on their academic performance. However, you might not be able to prove conclusively that the poor performance was caused by time used for cooking. Nevertheless, if you have carefully controlled other possible variables that might produce the difference, then a causal relationship exists. The use-of correlation research designs will enable you to map out the relationship between two or more educational variables.
CASE STUDY DESIGN
A case study seeks to describe a unit in detail, in context and holistically. It is a way of organizing educational data and looking at the object to be studied as a whole. In a case study, a great deal can be learned from a few examples of the phenomena under study. For example, The impact of pay-as-you-eat on education: A case study of Kenyatta University. A study of Kenyatta University can allow an in-depth investigation of the problem at hand. It will bring about deeper insights and better understanding of the problems faced by students. You should use case study design if you intend to analyze an issue in detail. Ensure that you have justified why you selected a case study.
CROSS CULTURAL RESEARCH DESIGN
This is used to compare the behaviour patterns of different cultures. Using this design you can perceive how various cultures perceive certain educational and social outcomes. For example, you can compare the performance of students in English in rural and urban schools and find out to what extent cultural variations influence performance.
Steps to Follow n Selecting a Research Design
Below are some of the steps a researcher should follow while selecting a research design:
- Identify the kind of research you intend to carry out. Being aware of the purpose and objectives of your study and your theoretical foundations will considerably influence how you design your research: where you go, for how long, with whom you talk, and the kind of questions you ask. Deciding if you intend to test or elaborate existing theory or are trying to build a new, grand theory, or are using existing theory in a new way, has implications in the kind of information you need to collect.
- Use the library to analyze samples of research designs from books and periodicals. The Internet is another option.
- Discuss with colleagues on the validity and reliability of your research and make a decision on what design will assist in answering your research questions appropriately.
Qualities of Effective Research Design
- They are systematic and logical. They effectively address the questions raised in the study. Based on this design the researcher can construct questions that will solicit the desired information.
- They contribute to accurate and fair interpretation of results.
- They clarify to the researcher the respondents and the means by which the study will be conducted-
- They contribute to deeper insights and better understanding of the research topic.
Guidelines in Selecting a Research Design
The following are essential points that researchers should adhere to while selecting a research design:
- Identify the research questions to be addressed by the study: The researcher should identify and reflect on the research questions raised in the study. Reflection should include brainstorming on issues such as:
- Do the questions raised in the study require systematic manipulation of independent and dependent variables? If the answer is yes, then the researcher will use an experimental design.
- Does the study require the researcher to assess the degree of relationship between two or more variables? If the answer is positive then a correlation design will be used.
- Does the study seek to describe a unit in detail? If so then a case study design will be used.
- Does the study seek to compare the behaviour patterns of different cultures? If the answer is positive then a cross-cultural researclti4esign will be applicable.
b) After identifying the research design to be used, read materials related to that design to understand its advantages and disadvantages. –
Indicate the research design pointing out its validity and reliability to the current research.
Pitfalls in Selection of Research Designs
While selecting a research design, a researcher should be on the lookout for the following pitfalls and avoid them:
- Choosing a design that cannot assist in meeting the research objectives.
- Choosing a design that is too complex for research at the level at which the student is studying.
- Choosing a design that requires extensive study and a lot of time while the time assigned to the research is limited.
- Lack of clarity about the design.
- A research design that lacks flexibility.
From the above, it is clear that in selecting a research design the researcher should ensure that it links concepts and questions with the study and it is specific and flexible and expansive enough to adapt to various complexities.
The selection of a research site is essential. It influences the. usefulness of the information produced. The idea is to start with larger population and through progressive elimination, end up with the actual site where data is collected (Orodho and Kombo, 2002). It is important to do the following:
- Identify the 1argt areas which are relevant to your research questions and objectives.
- Consider the heterogeneity of the potential study population and choose areas or communities which represent the range of variations on the most important characteristics.
- Identify and select actual communities which fulfill these criteria by making site visits, discussing with community leaders.
- Issues of accessibility should also be considered.
A population is a group of individuals, objects or items from which samples are taken for measurement (for example a population of students). Population refers to an entire group of persons or elements that have at least one thing in common, for instance, students at Kenyatta University). Population also refers to the larger group from which the sample is taken. It is important for the researcher to find out as much as possible about the study population. This includes some of the overall demographics such as age, gender and class of the population. The greater the diversity and differences that exist in the population, the larger the researcher’s sample size should be. Capturing the variability in population allows for more reliability of the study.
The following are qualities of an effective population sample:
- Diversity: An effective population sample attempts to be as diverse as possible. The greater the diversity and differences that exists in the population sample the higher the applicability of the research findings to the whole population.
- Representative: It is important for the researcher to identify and select respondents that fulfill the questions the research is addressing. For example, if a study is on the effect of the slum environment- of basic education, it is important that the majority of the population of the respondents is from the slum environment.
- Accessibility: An effective population sample is one that is accessible to the researcher.
- Knowledge: An effective population sample should have some idea of the topic being investigated.
Guidelines in Population
In population sampling, the researcher should carry out the following:
- Reflect on the research title particularly the independent and dependent variables and the study objectives. This enables the researcher to identify the type of population that will be most suitable for the study.
- Identify the largest population which can relevantly be used s respondents in addressing the research questions and meeting the specific objectives.
- Consider the heterogeneity of a potential study population and choose areas or communities which represent the range of variations with the most important characteristics.
- Evaluate the effectiveness of the selected population in meeting the objectives of the study. Issues of accessibility to the respondents should also be considered during evaluation.
- Sampling is the procedure a researcher uses to gather people, places or things to study. It is a process of selecting a number of individuals or objects from a population such that the selected group contains elements representative of the characteristics found in the entire group (Orodho and Kombo, 2002). A sample is a finite part of a statistical population whose properties are studied to gain information about the whole (Webster, 1985). When dealing with people, it can be defined as a set of respondents (people) selected from a larger population for the purpose of a survey. Research conclusions and generalizations are only as good as the sample they are based on. Samples are always subsets r small parts of the total number that could be studied. Sampling is the act process or technique of selecting a suitable sample, or a representative part of a population for the purpose of determining parameters or characteristics of the whole population. The way in which a researcher selects subjects for a study will determine how one is able to generalize the results of the study.
The term “sampling design” refers to that part f the research plan that indicates how cases are to be selected for observation. Sampling designs are divided into two broad areas:
- Probability designs.
- Non-probability designs.
The key component behind all probability sampling approaches is randomization, or random selection. In probability sampling people, places or things are randomly selected. Each unit in the population has an equal chance of being selected. This sampling gives every member of the population equal chances of being included, in the study. Probability sampling enables the researcher to generalize to the larger population and make inferences. If the purpose of your research is to draw conclusions or make predictions affecting the population as a whole, then probability sampling is appropriate Various methods have been established to accomplish probability sampling. These include the following:
a) Simple random sampling
This method is referred to as simple random sampling as no complexities are involved. All you need is a relatively small, clearly de- fined population to use this method. For example in a town of 10,000 residents, the researcher may simply obtain a-list of all residents, and then using a sequence of numbers from a random numbers table (or draws of a hat, flips of a coin), selects say 10% or 20%, or some portion of names on that list, making sure that he/she is not drawing from any letter of the alphabet more heavily than others. Advantages of simple random sampling are that the samples yield research data that can be generalized to a larger population. This method also permits the researcher to apply inferential statistics to the data and provides equal opportunity of selection for each element of the population. It is a procedure in which all the individuals in the defined population have an equal and independent chance of being selected as a member of the sample.
However this method also has disadvantages. This includes the following:
- It is not the most statistically efficient method of sampling. The researcher may, just because of luck of draw, not get good representation of subgroups in a population.
- Bias in selection is common.
- Some samples may be over or under represented.
- Non response error is high. Some of the members selected may have moved to other areas.
b) Stratified random sampling
Stratified random sampling involves dividing your population into homogeneous subgroups and then taking a simple random sample in each subgroup. The sample is selected in such a way as to ensure that certain subgroups in the population are represented in the sample in proportion to their number in the population. This method is appropriate when the researcher is interested in issues related to gender, race or age disparities in the population.
For example, if one is planning to study Factors influencing female enrollment in architecture and knows that gender is going to be an important factor because female students rarely take this course or quit before completing the course, the researcher therefore needs to stratify the sample by the gender strata, making sure that the female students are over sampled (draw more of random number of female students) as opposed to male students (which the researcher would under sample). For example, the department has 1,000 students consisting of 900 male and 100 female students, and the researcher’s intent on sampling 10% of the total, and that the researcher proceeds as usual, drawing 90 males at random and 10 females at random. If he/she had used the student list of names, regardless of gender, chances are that the researcher may not obtain 10 female students at random because they are fewer in total number. The advantages of this method are that you will be able to represent not only the overall population, but also key subgroups of the population, especially small minority groups. Stratified random sampling will generally have more statistical precision than simple random sampling.
If not carefully stratified, bias can occur resulting in some groups of the population being unrepresented
c) Systematic random sampling
Suppose a researcher had a large list of people, places or things to select from, such as 100,000 people or more. The appropriate method to use is to select every 10th, 20th, or 30th person from such a list. This decision to use every 10th , 20th , or 30th person is called the sampling interval, and as it is done systematically and the entire list is used, the researcher is said to be systematically random sampling.
- Large populations can be analyzed.
- Every member of the populations has an equal chance of inclusion.
- Bias is minimized.
- The response may be low since the respondents’ availability is unpredictable.
- The selection of the first sample member may result in a bias in the entire sample.
- The list used may not be in a systematic order.
d) Cluster random sampling
In the event that a population is dispersed across a wide geographic region, one may have to use cluster random sampling. This method allows for the division of the study population into clusters (usually counties, regions, provinces or other boundaries) and random sampling of everyone in those clusters. The units within the sampled clusters should be measured.
For instance, a survey of all secondary schools in Kenya will require the researcher to visit all the provinces. If one uses the simple random sampling method, he/she will have to cover the entire country geographically. Instead, one could simply do a cluster sampling of two districts per province, which would then be visited for the survey. The advantage of this method is that it needs a detailed sampling frame for selected clusters only rather than for the entire target area. There are savings in travel costs and time as well. However, there is a risk of missing important sub-groups and not having a complete representation of the target population.
Probability sampling is any method of sampling that utilizes some form of random selection. In order to have a random selection method, a researcher must set up some process or procedure that assures that the different units in the selected population have equal probabilities of being chosen. Some forms of random selection include picking a name out of a hat. These days, you can use a computer and generate random numbers as the basis for random selection. Random sampling is still regarded as one of the best statistical methods as it is free from bias.
- There is a risk of missing on important sub-groups
- Lack of complete representation of the target population
In this method, the researcher is interested in the representativeness of the concepts in their varying forms. This method of sampling aims to be theoretically representative of the study population by maximizing the scope or range of variation of the study. This method is mainly applied to find out how a small group, or a representative group, is doing for purposes of illustration or explanation. Various methods have also been established to accomplish non-probabilistic sampling.
a) Quota sampling
This sampling technique begins by dividing the population into relevant strata such as age, gender or geographical region. The total sample is allocated among the strata in direct proportion to their estimated or actual size in the population. Once the researcher identifies the people to be studied, they have to resort, to haphazard or accidental sampling because no effort is usually made to contact people who are difficult to reach in the quota. The problem with this method is that bias intrudes ‘On the sampling frame. This is because researchers allowed to self-select respondents are subject to bias such as interviewing their friends in excessive proportions or concentrating in areas where there are large numbers of potential respondents.
b) Convenience sampling
This method is based on using people who are a captive audience,. people the researcher meets haphazardly or accidentally. Respondents are people who just happen to be walking by, or show a special interest in your research. The use of volunteers is an example of convenience sampling.
c) Purposive sampling
In this sample method, the researcher purposely targets a group of people believed to be reliable for the study. For example, to study the effects of abortion on learning, the researcher may make efforts to contact students who previously had terminated their pregnancies. The researcher never knows if the sample is representative of the population. The power of purposive sampling lies in selecting information rich cases for in-depth analysis related to the central issues being studied.
Purposive sampling can be used with both quantitative and qualitative studies. Purposive sampling can be carried out in addition to probability sampling. For example, after completing your baseline study based on a random sample, you may recognize that certain sections of the project area are quite different from other areas due to variations in landscape, geography, culture etc. You may then purposively select those areas to get representative information about how the variations have influenced the behaviour of the people. Purposive sampling is particularly relevant when you are concerned with exploring the universe and understanding the audience. This means, using your common sense and the best judgment in choosing the right habitations and meeting the right number of the correct people for the purpose of your study. Types of purposive sampling include the following:
- Extreme Case Sampling: It focuses on cases that are rich in information because they are unusual or special in some way, for instance, the only community in a region that prohibits wife inheritance.
- Maximum Variation Sampling: Aims at capturing the central themes that cut across participant variations, for instance, persons of different age, gender, religion and marital status in an area protesting against child marriage.
- Homogeneous Sampling: Picks up a small sample with similar characteristics to describe some particular subgroup in depth, for example, charcoal burners, touts, bar maids, and so on.
- Typical Case Sampling: Uses one or more typical cases (individuals, families/households) to provide a local profile. The typical cases are carefully selected with the co-operation of the local people/ extension workers.
- Critical Case Sampling: Looks for critical cases that can make a point quite dramatically, for instance, farmers who have set up an unusually high yield record of a crop in arid lands.
- Snowball or Chain Sampling: Begins by asking people, “who knows a lot about “ By asking a number of people, you can identify specific kinds of cases, for example critical, typical, extreme and so on. Snowball sampling begins with a few people or cases and then gradually increases the sample size as new contacts are mentioned by the people you started out with.
- Purposive sampling is adequate under the following situations:
- When studying past events and only a fraction of relevant materials is available or accessible.
- While studying sensitive issues such as abortion, prostitution or crime, certain individuals or groups of individuals may refuse to cooperate. The researcher may use a non-probability method.
- If the population contains few relevant cases.
- If the population is unknown or not readily identifiable.
Target Population (Selection of Respondents)
The people a researcher selects as respondents in the study are vital in achieving the set objectives. Selection of respondents will largely depend on the following:
- Information needed
- Data techniques to be used
- The available funding may pre-specify the sample size.
For reliable conclusions to be drawn from the research, samples for quantitative research must be representative of the target group. Other things being equal, a larger sample of respondents is better than a smaller one. In general, the larger the sample, the more representative it is likely to be, and the more generalizable the results of the study are likely to be. Minimum acceptable sizes depend on the type of research.
Generally, a researcher would need 30 subjects in each group for co-relational and descriptive research but may be able to get by with 15 subjects per group in experimental or quasi-experimental designs. In general, selection of respondents will depend on the nature of the analysis to be performed, the desired precision of the estimates one wishes to achieve, the kind and number of comparisons that will be made, the number of variables that have to be examined simultaneously and how heterogeneously a universe is sampled. Population is a set of all the elements of interest in a study. Efforts should be made by a researcher to ensure that informants, particularly key informants, possess special knowledge related to the study. Efforts should be made to ensure the participants are active participants in the culture or organization under study, that they are involved in the events under study and have adequate time. They should be willing to talk to the researcher.
Bias and Error in Sampling
There are various challenges faced by researchers during sampling. Some of these challenges include the following:
Sampling error — Sampling error comprises of the differences between the sample and the population that are due solely to the particular units that happen to have been selected. For example, suppose that a sample of 100 university students is measured and all are found to be taller than six feet. It is very clear even without any statistical proof that this would be a highly unrepresentative sample leading to invalid conclusions. This is a very unlikely occurrence because naturally such rare cases are widely distributed among the population. But it can occur. Luckily, this is a very obvious error and can be detected very easily. The more dangerous error is the less obvious sampling error against which nature offers very little protection. An example would be a sample in which the average height is overstated by only one inch or two rather than one foot which is more obvious. It is the unobvious error that is of much concern.
There are two basic causes for sampling error; chance and sampling bias.
- Chance – This is the error that occurs due to bad luck. This may result in untypical choices. Unusual units in a population do exist and there is always a possibility that an abnormally large number of them will be chosen. The main protection against this kind of error is to use a large enough, sample.
- Sampling bias — Sampling bias is a tendency to favour the selection of units that have particular characteristics. Sampling bias is usually the result of a poor sampling plan. The most notable is the bias of non-response when for some reason some units have no chance of appearing in the sample. Take a hypothetical case where a survey is conducted to find out the level of stress that graduate students are going through. A mail questionnaire is sent to 100 randomly selected graduate students. Only 52 students respond. The results show that students are not under stress, yet the actual case is that stress levels may be high except among those who are answering the questionnaire. Bias can be very costly and has to be guarded against as much as possible. A means of selecting the units of analysis must be designed to avoid the more obvious forms of bias.
Non-sampling error (measurement error) The other main cause of unrepresentative samples is non-sampling error. Non-sampling error may either be produced by participants in the statistical study or may be an innocent by-product of the sampling plans and procedures. A non-sampling error is an error that results solely, from the manner in which the observations are made. The simplest example of non-sampling error is inaccurate measurements due to malfunctioning instruments or poor procedures. For example, consider the observation of human eights. If persons are asked to state their own weights themselves, no two answers will be of equal reliability. The people will have weighed themselves on different scales. An individual’s weight fluctuates, so that the time of weighing will affect the answer. The scale reading will also vary with the person’s state of undress. Responses therefore will not be of comparable validity unless all persons are weighed under the same circumstances. Biased observations due to inaccurate measurement can be innocent but very devastating.
In surveys of personal characteristics, unintended errors may result from the manner in which the response is elicited, the social desirability of the persons surveyed, the purpose of the study and the personal biases of the interviewer or survey writer. In all the sampling procedures the major weaknesses include failure to identify the accessible and target population and using a sample that is too small to permit statistical analysis.
Challenges Faced in Population Sampling
In population identification, researchers are sometimes faced with various challenges. These include the following:
- Scope: A very wide scope for example a study of the whole country may hinder effective sampling of the population. A narrow scope for example a study on one school affects the validity and reliability of the findings.
- Lack of representation.
- Bias in sampling: some researchers select a population that is convenient for them in terms of accessibility.
- Poor accessibility to the population: Some population samples are difficult to access.
In research, the term “respondents” refers to those who will reply to, or respond to the research instruments. The selection of respondents is crucial to the overall usefulness of the information produced. This is because respondents help in the clarification of issues under the study. This contributes to the achievement of set objectives. The selection of respondents will largely depend on the information needed and the date techniques to be used. The researcher should ensure that informants, particularly key informants, possess special knowledge related to the study area.
Qualities of Effective Respondent Selection
The following should be adhered to by researchers in the selection, of respondents:
- Respondents should be individuals who possess some knowledge about the topic being studied.
- They should be willing to share the information they have in relation to the topic with the researcher.
- They should be active participants in the culture or organization under study.
- They must be willing to give their time to the study.
- A large sample of respondents is better than a small one. In general, the larger the sample, the more representative it is likely to be, and the more generalizable the results of the study are likely to be.
Challenges Faced in Respondent Selection
The selection of reliable informants has various challenges. These include the following:
- Unwillingness of respondents to share all they know on the issue with the researcher.
- Language barrier: the interview or questionnaire may have been written in Kiswahili yet the respondent can effectively express him! herself in English or, say, Dholuo.
- Hostility towards the researcher: some respondents may personalize the questions asked particularly during interviews and become hostile towards the researcher.
- Time limitations.
In general, selection of respondents will depend on the nature of the analysis to be performed, the desired precision of the estimates one wishes to achieve, the kind and number of comparisons that will be made, the number of variables that have to be examined simultaneously and how heterogeneously a universe is ‘sampled.
Research instruments include the following: questionnaires, interview schedules, observation and focus group discussions.
In formulating research instruments the researcher should ensure the following:
- The objectives of the study are clear. This will assist the researcher to anticipate the type of information needed.
- The population sample. The researcher should be aware that some types of instruments are unsuitable to some groups of people due to factors such as the literacy level, profession and culture. A researcher should determine the literacy level of the study population in advance. For the illiterate, interview and focus group discussions should be used. The type of language that will be used (either English or Kiswahili) will depend on the literacy level of respondents.
- Geographical distribution. The span of the study dictates the type of instrument to be used. A countrywide study may require the use of postal questionnaires and telephone interviews.
- A researcher should be careful about the questions he/she asks. According to Orodho and Kombo (2002), a researcher should do the following::
- Begin with a few interesting but non-threatening questions.
- Avoid vague questions, for instance, “What do you like?”
- Keep the language simple.
- Limit each question to a single idea. Ensure each item included has a specific purpose.
- Only include questions that are directly relevant to the study.
- Have a logical sequence.
- Do not put the key questions at the end of the questionnaire. It is best to have them in the middle.
- Avoid emotionally charged words.
- Avoid leading questions, for example, “Do you think students riot because they are unfairly treated?”
- Avoid acronyms ‘and abbreviations.
- Consider the order of questions related to each topic.
- A researcher should vary closed and open ended questions. Closed questions give the respondent a set of choice or options. Open-ended questions are free response type questions. They allow the respondent to answer in their own words.
- Check the consistency of answers. It may be beneficial to ask the same question again using different wording. This ensures validity.
This is a research instrument that gathers data over a large sample Questionnaires have various advantages including the following:
- Information can be collected from a large sample and diverse regions.
- Confidentiality is upheld.
- Saves on time.
- Since they are presented in paper format there is no opportunity for interviewer bias.
However they have their disadvantages in that:
- Response rates can be quite low. –
- There is no direct contact so the researcher cannot deal with any misunderstanding.
- There is no opportunity to ask for further information related to answers given.
- No clear reason can be given for incomplete responses.
- To ensure the effectiveness of questionnaires a pre-test should be carried out. Pilot the questionnaire with a small representative sample. This will enable the researcher to find out if:
- The questions are measuring what they are supposed to measure.
- The wording is clear.
- If all questions are interpreted in the same way by respondents.
- What response is provoked.
- If there is any research bias.
Steps in Formulating a Questionnaire
The following are essential in the formulation of a questionnaire:
- Reflection — Before formulating a questionnaire it is important for the researcher to reflect on the aim and objective of the study. The researcher should reflect on the type of response expected.
- Formulation of questions — The researcher should write down questions related to each stated objective. While constructing the questions the researcher should begin with a few interesting but non- threatening questions. The researcher should only include questions that are relevant to his/her study. The researcher should keep the key questions in the middle.
- Pilot — After constructing the questionnaire, the researcher should try it out on a small sample of the population. While piloting the researcher should address the following questions:
- Are the questions measuring what they are supposed to measure — the researcher should analyze each answer and see if it is supplying the appropriate information.
- Is the wording clear? The researcher should analyze the responses to find out if there was any confusion in the way questions were interpreted by all the respondents.
- Do the questions provoke a response? If some questions have been omitted, the researcher should find out why.
- Is there researcher bias? The researcher will analyze whether the questions asked were skewed towards certain issues more than others.
- Evaluation — After piloting and making the necessary amendments, the researcher should carry out an evaluation of the revised questions. This includes finding out if the questions are clear and specific, where the key questions are placed and if the balance of questions is correct.
Qualities of an Effective Questionnaire
An effective questionnaire has the following qualities:
- It is simple to understand. The language used is clear and straight forward. This helps reduce misinterpretation.
- Instructions are clearly given. There are a few words of explanation in each new section.
- The questions are focused and are limited to a single idea. Sentences are short and precise.
- Each item included has a specific purpose, and contributes to the study.
- There are no leading questions. ,,‘
- There is a balance of questions per topic.
- Can cover a wide area
- No bias on the side of the researcher and the respondents
Questionnaires have the following disadvantages
- The response rate can be quite low. Since the researcher is not in direct contact with the respondents they may not feel the obligation to complete the questionnaire as soon as possible. This postponement in completion can result in the questionnaire not being answered at all.
- There are no direct contacts between the researcher and respondent. The researcher cannot therefore deal with or clarify any misunderstanding.
- There is no opportunity for the researcher to ask for further information, or probe deeper into answers given by the respondent.
- Incase some questions are not answered, the researcher cannot get an explanation from the respondent as to why some questions are incomplete.
- The researcher is not able to predict if respondents have answered all the questions until after the collection of the instrument.
- The researcher has no control over the order in which questions are answered. Yet in research, the way questions are answered can pre-determine their validity.
These are questions asked orally. There are various forms of inter- as follows:
In this approach to interviewing, the researcher has some idea in mind of the topics to be covered and may use some sort of topic list as a reminder. There is minimal control over the order in which topics are covered and over respondents’ answers. In unstructured interviewing, neither the specific questions to be asked nor the range or type of possible answers are pre-defined. They are informal and conversational. The aim is to get the informants to open up, and the researcher should stimulate an informant to produce more information.
This approach allows the interviewer to be responsive to individual differences and situational characteristics. This approach builds on observation. It is useful in studying sensitive topics such as sexuality or political topics. To effectively achieve the aims of an open-ended interview, one must ask a whole series of secondary questions such as:
- What do I want to get out of these interviews?
- With whom am I going to conduct these interviews?
- How do I know they will talk to me?
- How many interviews must I do?
As much as possible, test your methods in advance. Because there is no set format for conducting these interviews, each interview is unique. This makes it difficult to systematize and analyze data.
Unstructured interviews have the following advantages:
- They are flexible. This is because there are no pre-defined questions. This allows the respondents to freely respond to an issue. The researcher can therefore gather a lot of information.
- The respondent feels part of the team since no rigidity is displayed. He/she can therefore freely participate in the research.
- Since it is a free response in a relaxed atmosphere situation, the answer given are more reliable.
- It allows the interviewer to be responsive to individual differences and situational characteristics.
- This instrument is useful for studying sensitive topics such as sexuality and abortion.
Unstructured interviews have the following disadvantages:
- They are time consuming since a respondent can dwell on one issue.
- They are not systematic as a respondent can comment on issues in a haphazard way. A respondent can comment on issues already discussed.
- If the researcher is not careful, it can get out of control, with the respondent getting too emotional or personal.
- Irrelevancies can be displayed by the respondent.
- Since there is no set format for conducting these interviews, it is difficult to systematize and analyze data.
These interviews are based or the use of an interview guide. This is a written list of questions or topics that need to be covered by the interview. There are several types of semi-structured interviews.
- Focused interviews — This intensively investigates a particular topic. They aim at gaining a complete and detailed understanding of the topic.
- Case studies — The purpose of case studies is to collect comprehensive, systematic and in-depth information about particular cases of interest.
Semi-structured interviews have the following advantages:
- They are flexible. This is because they consist of both open and closed-ended questions.
- In-depth information is gathered by closed ended questions.
- By using both the open and closed-ended approach, the researcher gets a complete and detailed understanding of the issue under research.
Semi-structured interviews have the following disadvantages:
- They can be time consuming due to the open-ended questions.
- Analysis of data may be problematic.
- The respondent may be cautious of the answers given in close -ended questions.
These involve subjecting every informant in a sample to the same stimuli, for instance, asking each informant similar questions, as in the case of a survey.
Structured interviews have the following advantages:
- The reliability of the information gathered is high. This is because each informant is subjected to similar questions with the others.
- It gives in-depth information about particular cases of interest to the researcher. This is because the researcher seeks information on specific issues.
- It is systematic. Researchers intensively investigate a particular issue before moving to the next.
- It is time-saving since the respondents simply answer what has been asked by the researcher.
- The researcher gets a complete and detailed understanding of the issue from the respondent.
- It is comprehensive and systematic since questions are formulated before the interview.
- The data collected is quantifiable.
Structured interviews portray the following disadvantages:
- The rigidity displayed by the researcher can affect the responses given. The respondent may feel as if he/she is under investigation and is being probed. This may affect the response. Some of the respondents may become hostile.
- it is too formal. Since the researcher does the questioning and the respondent simply answers, the respondent may be too cautious in the answer given. The respondent may give answers he/she thinks are acceptable or will impress the researcher.
- The researcher may miss out on some important points that are not included in the questions formulated.
Focus Group Discussions
This is a special type of group in terms of its purpose, size, composition and procedures. A focus group is usually composed of 6-8 individuals who share certain characteristics, which are relevant for the study. The discussion is carefully planned and designed to obtain information on the participants’ beliefs and perceptions on a defined area of interest. Special predetermined criteria are used in selecting focus group participants. This includes the following:
- The topics to be discussed are decided beforehand.
- There is a predetermined list of open ended questions.
- Focus relies on discussion among participants about the topics presented.
This method requires thorough planning and training of group moderators. Focus groups should usually be composed of homogeneous members of the target population, for instance, similar in age, education level, gender, profession. Focus group discussions can produce a lot of information quickly and are good for identifying and exploring beliefs, ideas or Opinions in a community. However, the
researcher has less control over the ‘flow of the discussion and results are hard to analyze. Focus group discussions are used to assess needs, develop intervention, test new ideas or programmes or improve existing programmes.
This is a tool that provides information about actual behaviour. Direct observation is useful because some behaviour involves habitual routines of which people are hardly aware. Direct observation allows the researcher to put behaviour in context and thereby understand it better. Observation can be made of actual behaviour patterns. Forms of observation include the following:
a) PARTICIPANT OBSERVATION
The investigator becomes an active functioning member of the culture under study. An investigator participates in any activity appropriate to the status which is assumed. This participation helps reduce reactivity. Respondents become more comfortable with the researcher. It gives a researcher an intuitive understanding of what is happening in a culture. However, it can be time consuming.
b) UNSTRUCTURED OBSERVATION
The observer takes the position of an onlooker. Data is collected in the form of descriptive accounts. Unstructured observations are helpful in understanding behaviour patterns in their physical and social context.
c) STRUCTURED OBSERVATION
The observer is an onlooker. The focus is on a small number of specific behaviour patterns, and only those appearing on a pre-defined observation list are recorded. This requires the researcher to be clear on the behaviour being observed.
Standardized tests of one sort or another are used in most educational research studies. A researcher will frequently use standardized tests to measure one or more of the variables in a study. It is important that one gets as much information as possible about the tests to be used in the study. in some cases no suitable instrument exists to measure the variables of the study. In that case, the researcher will have to design their own instrument. One can look at the instruments (such as questionnaire forms), which have been used in similar studies and modify these for use in his/her own study. There are many different types of tests that one might consider for use in their study. Some of the most commonly used types of tests educational research are:
- Achievement tests
- Personality tests
- Aptitude tests, including tests of academic aptitude (intelligence tests)
CHARACTERISTICS OF STANDARDIZED TESTS
Validity – the validity of a test is a measure of how well a test measures what it is supposed to measure. The examiner’s manual or technical manual for most tests will have information on the validity of the test.
Reliability — reliability is a measure of how consistent the results from a test are. If you administer a test to a subject twice do you get the same score on the second administration as you did on the first? The reliability of the test is the answer to this question.
Validity and Reliability of Research Instruments
No two interviewers are alike and the same person may provide different answers to different interviewers. The manner in which a question is formulated can also result in inaccurate responses. Individuals tend to provide false answers to. particular questions. For example, some people want, to feel younger or older for reasons known to themselves. If a researcher asks such a person their age in years, it is easier for the individual just to lie by overstating their age by one or more years than it is if the researcher asked which year they were born since it will require a bit of quick arithmetic to give a false date. A date of birth will definitely be more accurate.
The respondent effect. Respondents might also give incorrect answers to impress the interviewer. This type of error is the most difficult to prevent because it results from outright deceit on the part of the respondent. For example in asking farmers how much maize they harvested, the farmers may lie by saying a figure which is the recommended expected yield that is 25 bags per acre. The responses may therefore appear uniform. The researcher should be suspicious and can compared this with the responses of the farmers’ spouses. To decide which answer is most accurate, whenever possible the researcher should in a tactful way verify with an older son or daughter. It is important to acknowledge that certain psychological factors induce incorrect responses and great care must be taken to design a study that minimizes this effect.
Data Collection Procedure
A researcher will require a research permit before embarking on the study. The researcher will then administer research instruments to the respondents.
Data collection must be accurate. Where, tests are used they must be scored correctly, and observations must be made systematically. In some cases data may be coded, for example males coded as 1 and females coded as 2. An electronic spreadsheet is an excellent place for the researcher to keep the data for the study. This includes both raw data and coded data. In most cases you will also be able to perform the desired statistical calculations from within the spreadsheet. The MS Excel spreadsheet programme, for example, has an Analysis Tool Pack that will allow one to calculate such statistics as chi-square, correlation coefficient, t-test, z-test, and analysis of variance. The major ways of collecting data include administering a standardized instrument, administering a self—developed instrument and recording of naturally available data.