Here is the link for part 1
Here is the link for part 2
CHAPTER THREE
METHODOLOGY
Chapter Introduction As earlier stated, it is good practice to have in the first section of a chapter (apart from chapter one which introduces the problem statement) an introduction to the chapter. The chapter introduction should show what the chapter is about (what it seeks to achieve) and how the chapter is structured into subsections.
Data Requirement It is reasonable to indicate early in this chapter the type of data that would be required to fulfil the objectives and answer the research questions. For example, if the objective is in portfolio management, to analyse the return of property and stock market assets, then the data requirement might be rental value and capital value data for say ten years (for property assets) and dividend data and share price data for the same period of time.
The data requirement above would assist in pointing out whether the data requirement would be sourced through secondary data or primary data or both. In the portfolio management example above, the stock market data would be obtained from secondary data published on the internet, while the property market data would be obtained from (primary data) surveys of estate surveyors and valuers.
Study Population Where the data is to be sourced from primary sources, the statement of the data required (above) would provide the opportunity of identifying which people or buildings or items have the characteristics would be appropriate to provide primary data. If a researcher is interested in capturing residential rental value at increasing spatial gradients from a business district for example, then the residents located at the within each of the spatial gradients would be an appropriate study population as well as the estate surveyors and valuers in the study area
Sample Frame the sample frame is the total number of population elements in an authentic list of each of the population units in the study population. It is important that the researcher identify and state where he obtained such an authentic list of the population. Where the study population is estate surveyors and valuers in a state or city, an authentic list of population elements could be obtained from the most recent Directory of Estate Surveyors and Valuers. Where this is outdated, it could be supplemented with lists from the state branch of the Nigerian Institution of Estate Surveyors and Valuers, especially since the Directory lists refer only to head offices.
Sample Size – Generally, the larger the sample size, the more accurately the sample will reflect the population it was drawn from. There are formulae for calculating the minimum number of sample that would adequately represent the population.
Method of Sampling There are two types of sampling a survey researcher may employ: probability sampling and non-probability (judgemental) sampling. A probability sample is a sample in which every element in the population has a chance of being selected in the sample and the probability can be accurately determined. Probability samples include
Simple random sampling,
Systematic sampling the population is arranged according to some ordering scheme (e.g. alphabetically) and every kth element is selected at regular intervals where k = population size/sample size. The starting point is randomized, making systematic sampling a form of probabilistic sampling..
Stratified sampling the sample frame is organized into mutually exclusive categories according or strata and individual elements are randomly selected within each stratum.
Cluster sampling this method is usually undertaken to reduce transport and administrative costs. It involves identifying a cluster of elements (clustered either geographically or by time frame) in a sample frame and interviewing all elements within the cluster.
Multistage sampling this involves taking a process of taking random sub-samples of preceeding random samples. In the first stage, clusters are constructed. In the second stage, a sample of primary units is randomly selected from each cluster rather than using all elements in the cluster. In the subsequent stages, samples are selected from the primary units and so on.
These ways of sampling have the following properties: each element has a known probability of being sampled and the sampling involves random selection at some point.
Non-probability sampling is any sampling method where some elements of the population have no chance of selection or where the probability of selection cannot be accurately determined. These conditions give rise to exclusion bias placing limits on how much information a sample can provide about the population. Non-probaility sampling methods include
Accidental sampling (convenience sampling) – Here the sample is drawn from the part of the population that is close at hand (readily available and convenient). However, the researcher using such a sample cannot make generalizations about the entire population.
Quota sampling Here the population is first segmented into mutually exclusive sub groups as in stratified sampling. Then judgement is used to select the subjects or units from each segment based on a specified proportion for example a quota of 200 female 300 male estate surveyors.
Purposive sampling.- Here the researcher chooses the sample based on who they think would be appropriate for the study. It is used primarily where there is a limited number of people that have expertise in the area being researched.
Snowball sampling – Here the first respondent refers a friend. The friend also refers a friend and so on. Such samples are biased because they give people with more social connections a higher chance of selection.
Data Collection Instruments There are various methods of data collection
Experiments a controlled study in which the researcher attempts to understand cause and effect relationships. This is more often used in research in the physical sciences.
Observational studies – the researcher is as a participant observer in the subject matter
Questionnaire surveys The variants of this include self-administered questionnaire, telephone questionnaire and web-based questionnaire etc
Interview surveys The variants of this include face-to-face interviews, telephone interviews etc
Focus group discussions a planned discussion between 7 to 12 selected people (selected because they have certain characteristics that are relevant to the subject matter of discussion) coordinated by a skilled interviewer to share ideas and perceptions on the subject matter
Methods of Measuring Data – In the questionnaire, data can be measured using any of 4 scales of measurement.
Nominal scale This is the lowest level of measurement. One simply measures non-numeric responses in categories with labels such as male or female. There is no sense in which one category is placed ahead of the other.
Ordinal (Likert) scale This also measures non-numeric concepts in categories that are ordered e.g. strongly agree, agree, disagree, strongly disagree etc. It is used to measure non-numeric concepts as satisfaction of tenants with property management etc.
Interval scale This is a numerical scale of measurement where the categories are not just ordered, but difference between the categories is known. For example, the difference between time is known, consistent and measureable. Likewise, the difference between temperature on the on the Celsius scale. However, in the interval scale, there is no true origin in zero. There is no such thing as zero temperature.
Ratio scale This is a numeric scale that is ordered, shows the difference between variables and additionally has an origin in zero. It therefore allows for a wide range of descriptive and inferential statistics. An example is the measurement of rental values.
A parametric test is a test that measures data using interval or ratio scales and which assumes populations or sampling distributions with normal distributions. A non parametric test is a test that measures data using nominal or ordinal scales and which does not make assumptions about the distribution of the population,.
Method of Analyzing Data Data can be analyzed using descriptive or inferential statistics. Descriptive statistics describe the data (e.g. frequency counts, measures of dispersion such as standard deviation, measures of central tendency such as means) while inferential statistics make inferences about the data (e.g chi square test, t-test, f- test, z-test). An inferential test is used with data that is normally distributed.
At the end of this sub-section, it may be good practice to present a table which captures information on data requirement, data measurement and data analysis methods for each objective as follows
Objective
Data requirement
Data measurement
Data analysis
Operationalization of variables
Pilot Survey A trial pre-run of the research survey to test validity and reliability of the design and administration of the research instrument
Validity & Reliability Tests validity refers to the degree to which a measuring instrument is correctly designed and administered. Reliability refers to consistency between independent measurements of the same phenomenon.