The Centre for Advanced Procurement Studies carries out empirical research in the areas of procurement, supply chain, and contracts management. The purpose of the research studies is to inform public policy and private sector strategy. The studies are carried out in Kenya and other African countries for informed solutions to continental challenges in a global marketplace.


Research designs and reports in this site are based on the APA Style accessible from www.apastyle.org


This section presents some of the basic terms used in empirical research and statistics.

Variable. A variable is a measurable characteristic that varies. It may change from group to group, person-to-person, or even within one person over time. These characteristics may be gender, education level, or procurement time.

Independent variable is the variable that is varied or manipulated by the researcher. Dependent variable is the variable that is being measured in a research. Your research might be to determine the effect of procurement certification on projects success. The researcher would like to know if those certified in procurement report higher level of success in procured projects than those who have no certification. Hence, the independent variable is certification and the dependent variable is project success.

Scale. A measurement scale is used to categorise and quantify variables. In a nominal scale, values assigned to the variable represent a descriptive category and have no inherent numerical value. Persons may be classified as male or female. Each value in an ordinal scale has a unique meaning and an ordered relationship to every other value on the scale. Education level is measured on an ordinal scale of (1) certificate, (2) diploma, and (3) degree signifying an order in the level of education achieved. Interval scale. Interval scale provides info about order and also possess equal intervals such as on a scale of 1 to 5. Ratio scale, in addition to possessing qualities of nominal, ordinal, and interval scales the ratio scale has an absolute zero. Examples of ratio scale are weight or price of an item.

Population. A complete set of persons or objects that possess some common characteristic. A population (N) could be all public procurement professionals in Kenya. It could also be all procurement transactions in a fiscal year. Accessible population is the portion of the population to which the researcher has reasonable access.

SampleThe selected elements of people or objects that are chosen to participate in a research. The sample should be representative of the population from which it was drawn. A sample represents characteristics of the population so that the findings from the sample can be generalised to the population. Sampling is the process of selecting subjects for a study in such a way that they represent the larger group from where they were selected. Sampling error is reduced by randomisation where every member of the population is equally likely to be selected into the sample. The number of subjects in the sample is the sample size (n).

Unit of analysis. These are the individual people, organisations, public entities that are under study. In most researches the unit of analysis will be an individual: procurement staffs, accounting officers, contracts specialists, procurement transactions, stores quantities, or number of transport shipments. When a research is to examine something that is not an individual, the researcher must decide who can most accurately report on the unit of analysis.

Alpha. Alpha is threshold value used to judge whether a test statistic is statistically significant. It can be denoted as α = 0.05 but can be adjusted to 0.01 for greater precision. Alpha is the probability of rejecting the null hypothesis when it is in-fact true. Alpha can range between 0 and 1.

Statistical power. The power or sensitivity of a statistical test is the probability that it correctly rejects the null hypothesis when it is false. It is the odds of saying that there is a relationship or price difference when in fact there is one. It is the odds of confirming our theory is correct. Power analysis is used to calculate the minimum sample size or calculate the minimum effect size that is likely to be detected in a study using a given sample size.

StatisticA statistic is a numerical value and measure of a characteristic of the sample. The mean, median, and mode are examples of statistical averages. The mean is the average where you add up all the numbers and then divide by the number of numbers. When numbers are listed in numerical order, the median is the middle value in the list of numbers. A mode is the value that occurs most often. A range is a measure of dispersion and is the difference between the largest and smallest values.

Standard deviation (sigma, σ ) is a measure of dispersion that is used to quantify the amount of dispersion of a set of data values. The variance (σ2 ) gives a measure of how the data distributes itself about the mean or expected value. Variance is calculated by taking the differences between each number in the set and the mean, squaring the differences and dividing the sum of the squares by the number of values in the set.

PrecisionAccuracy refers to how close a measurement is to the true value. Precision is how much variation there is when making repeated measurements. To evaluate the precision of data, the common statistics to use are the mean and the standard deviation. A large standard deviation means the measurements are far apart and imprecise.

Effect size. In statistical inference, an effect size is a measure of the strength of the relationship between two variables. An effect size is a useful descriptive statistic because it enables comparison across studies and hence is critical in meta-analysis studies. The most popular effect size is Cohen’s d, which is expressed as d, where M is the sample mean, μ is the population mean, and σ is the population standard deviation.

Correlation. Correlation is the measure of the relationship between two or more variables. Correlation analysis determines the strength and direction of the relationship between variables. A value of -1 means a perfect negative correlation while a value of +1 is a perfect positive correlation. A value of 0 means no correlation. A widely used correlation coefficient is Pearson r. If the correlation coefficient is squared, the resultant value r2 is called a coefficient of determination, which means the strength of the relationship between the variables.

Experiment. In an experiment, the researcher manipulates one or more variables, while holding other variables constant. An experimental group is the group receives the variable being tested. The control group does not receive the test variable.

Research design. The research design describes the study approach such as descriptive, correlational, semiexperimental, experimental, literature review, and meta-analysis. The design could also be descriptive, cross sectional, longitudinal, or case study. The design outlines the research question, hypotheses, independent and dependent variables, data collection methods, analysis of data, and reporting.

Normal distribution. A normal distribution has a mean equal median equal mode, a symmetry about the centre, and 50% of values are less than the mean and 50% greater than the mean (Figure 1). The normal distribution formula is:

Figure 1. Normal distributions of variable x

Figure 1. Normal distributions of variable x

Parametric tests. Parametric tests are based on assumptions about the normal distribution of the underlying population from which the sample was drawn. Nonparametric tests do not rely on assumptions about the shape or parameters of the underlying population distribution. The parametric assumption of normality is particularly an issue for small sample sizes (n < 30). Nonparametric tests are often a good option for these data.

Factor analysis. Is used mostly for data reduction purposes from a large set of variables to create indices with variables that measure similar things. The variables such as academic qualification, procurement certification, short term training, and professional membership may be used to measure one factor of professionalism. The key concept of factor analysis is that multiple observed variables have similar patterns of responses because they are all associated with a latent variable that is not directly measured.

Descriptive statistics. Descriptive statistics are used to describe the basic features of the data in a study. They provide summaries in numbers about the measures in the sample. When used in graphical form, they form a basis for most quantitative analysis in data. Descriptive statistics is used simply to describe what is going on in the data.

Inferential statistics. The researcher tries to reach conclusions that extend beyond the current data alone. It is used to infer from the sample what the population would be like. Inferential statistics is used to make inferences from the data to more general conditions of the population.

Hypothesis testing. A hypothesis is a statement about the relationship between two or more variables, which predicts the outcome of the study. It is an idea or explanation that could be tested through study. A hypotheses guide us on the selection of a certain design, observation, and method of researching a topic over others. Hypothesis testing is a method to test a claim or hypothesis about a parameter in a population by using sample data. The steps are: state the hypotheses, set the criteria for a decision, compute the test statistic, and make a decision. The null hypothesis (H0) is a statement about a population parameter that is assumed to be true. We test if the value stated in the null hypothesis is likely to be true. A researcher might state the null hypothesis as: There is no significant difference in bid prices for goods procured through open tender and restricted tender. An alternative hypothesis (H1) is a statement that directly contradicts a null hypothesis and states that the actual value of a population parameter is less than, greater than, or not equal to the value stated in the null hypothesis. H1 in this example would be: There is a significant difference in bid prices for goods procured through open tender and restricted tender. Significance level refers to a criterion of judgment upon which a decision is made regarding the value stated in a null hypothesis. The criterion is based on the probability of obtaining a statistic measured in a sample if the value stated in the null hypothesis were true. In social research, the level of significance is set at 5% (.05) but in natural sciences it could be 1% (.01).

P-value. A p-value is the probability of obtaining a sample outcome, given that the value stated in the null hypothesis
is true. The p-value for obtaining a sample outcome is compared to the level of significance. When the p-value is less
than 5% (p< .05), we reject the null hypothesis.

Regression analysis. Regression analysis is used to estimate if two or more variables are connected in a linear relationship. In simple regression, there are two variables being related. The formula for simple linear regression in a population is y = β0 +β1x +ε where y is the dependent variable, β0 is the intercept or constant, β1 is the xcoefficient that is the slope of the straight line of the equation, and ε is the error.

Research Methods

Research Method

Description of the research method

Descriptive Qualitative

Is a detailed description of a specific aspect of procurement, supply, or contract management using interviews, observations, and review of documents. The author describes things as they are. Such as how do the staffs procure infrastructure works? Which people make decisions on
contract awards?

Descriptive Qualitative

This is numerical description and measurement of things as they are such as in a survey. It involves measures of frequencies, averages, and numbers. The researcher might be interested to know the number of professionally qualified procurement staff in the organisation. Or, what proportion of total procurement volume was done through open competitive bidding?

Correlational Analysis

A quantitative analysis of the strength of the relationship between two or more variables. Is there a relationship between procurement certification and successful projects? Is restricted tender method associated with lost value for money?

Quasi Experiment


Compares a group that received a particular treatment and another group that is similar characteristics but did not receive the treatment. There is no random assignment of subjects to the two groups. Example, were the staffs that attended the CAPS practical training programs better in their job than those who did not receive the training? Did those that qualified in the CPSP(K) program the best performing in their job compared to the non-certified?


Uses random assignment of subjects to a treatment group and a control group where one receives the treatment and another does not. In a training workshop, you could randomly assign participants to two groups. Group A is trained using powerpoint presentations without input from the participants. Group B receives practical training based exercises and group discussions coupled with plenary discussions. The participants from the two groups never meet or communicate with each other. At the end of the 5-day training, you subject both groups to the same set of multiple choice questions on the materials covered during the two trainings. Is there any significant difference in test scores

Meta Analysis

Meta analysis is the synthesis of results from several multiple studies. The aim is to determine the impact from similar studies and form an opinion on the general trend in the area.

Research Articles Available For Download

  1. Impact of Procurement Practices on Projects Success in East Africa
  2. An Insight Into Procurement Standard Practices in Africa
  3. The Status of the Procurement Profession in Kenya Baseline Indicators 2011
  4. Procurement Value Chain Analysis

Fill in Your Name and E-mail to Get Download Link