Professional Documents
Culture Documents
By PAMELAH N. KIHEMBO
1. Definitionsh
Statistics: The term statistics is used to mean either statistical data or statistical methods.
The field of statistics: The study and use of theory and methods for the analysis of data arising from random
processes or phenomena. Or it is the study of how we make sense of data.
The field of statistics provides some of the most fundamental tools and techniques of the scientific methods
forming hypotheses
designing experiments and observational studies
gathering data
summarizing data
drawing inferences from data e.g. testing hypotheses
A statistic rather than the field of statistics also refers to a numerical quantity computed from sample data
e.g. the mean the median the maximum.
Biostatistics is the branch of applied statistics directed toward applications in the health sciences and biology.
Biostatistics is sometimes distinguished from the field of biometry based upon whether
applications are in the health sciences (biostatistics or in broader biology (biometry e.g
agriculture, ecology and wildlife biology)
Other branches of applied statistics: psychometrics, econometrics, chemometrics, astrostatistics,
environmetrics etc.
Because some statistical methods are more heavily used in health applications than elsewhere e.g
survival analysis, longitudinal data analysis)
Because examples are drawn from health sciences, biostatistics
1
Statistical data: When it means statistical data it refers to numerical descriptions of things. These descriptions
may take the form of counts or measurements.
NB Even though statistical data always denote figures (numerical descriptions) it must be remembered that all
'numerical descriptions' are not statistical data.
Statistical methods: When the term 'statistics' is used to mean 'statistical methods' it refers to a body of
methods that are used for collecting, organising, analyzing and interpreting numerical data for understanding a
phenomenon or making wise decisions. In this sense it is a branch of scientific method and helps us to know in
a better way the object under study.
1. Statistics pervades a way of organizing information on a wider and more formal basis than relying on the
exchange of anecdotes and personal experience.
2. There is a great deal of intrinsic (inherent) variation in most biological processes
3. Public health and medicine are becoming increasingly quantitative. As technology progresses, we
encounter more and more quantitative rather than descriptive information. In one sense, statistics is the
language of assembling and handling quantitative material.
4. The planning, conduct, and interpretation of much of medical research are becoming increasingly reliant
on statistical technology. For example answers to questions like:
Is this new drug or procedure better than the one commonly in use?
How much better?
What, if any, are the risks of side effects associated with its use?
In testing a new drug how many patients must be treated, and in what manner, in order to
demonstrate its worth?
What is the normal variation in some clinical measurement?
How reliable and valid is the measurement?
What is the magnitude and effect of laboratory and technical error?
How does one interpret abnormal values?
5. Statistics pervades the medical literature. As a consequence of the increasingly quantitative nature of
public health and medicine and its reliance on statistical methodology, the medical literature is replete
with reports in which statistical techniques are used extensively.
Limitations of statistics:
1. It deals with only those subjects of inquiry that are capable of being quantitatively measured and
numerically expressed.
2. It deals on aggregates of facts and no importance is attached to individual items–suited only if their group
characteristics are desired to be studied.
3. Statistical data are only approximately and not mathematically correct.
2
3. TYPES OF DATA
Data are observations of random variables made on the elements of a population or sample.
Data are the quantities (numbers) or qualities (attributes) measured or observed that are to be collected and
or analyzed.
Any aspect of an individual that is measured and take any value for different individuals or cases, like blood
pressure, or records, like age, sex is called a variable.
It is helpful to divide variables into different types : The main division is into qualitative (or categorical) or
quantitative (or numerical variables).
QUANTITATIVE/NUMERICAL VARIABLES
Quantitative variable: A quantitative variable is one that can be measured and expressed numerically and they
can be of two types (discrete or continuous).
Qualitative variables
A variable or characteristic which cannot be measured in quantitative form but can only be identified by name
or categories, for instance place of birth, ethnic group, type of drug, stages of breast cancer (I, II, III, or IV),
degree of pain (minimal, moderate, severe or unbearable).
In the above situation, we only know that the data are ordered.
3
Qualitative Nominal data:-
• Have no ranked order
– each category is no better/worse than other categories
– the gap between categories has no meaning
• Taking average of the categories is meaningless
• E.g.
– Sex (Male / Female)
– Ethnicity (White / Black / Asian / Chinese / Other)
– Blood group (A / B / AB / O)
– Eye color - brown, black, etc.
– Religion - Christianity, Islam, Hinduism, etc
CLASS EXERCISE-Assignment 2
Identify the type of data represented by each of the following examples.
1. Blood group
2. Temperature (Celsius)
3. Ethnic group
4. Job satisfaction index (1-5)
5. Number of heart attacks
6. Number of accidents in 3 - year period
7. Number of cases of disease reported by a health worker
4
TOPIC 2a: STUDY DESIGNS
Introduction
Studies can be classified as either observational or experimental.
1. EXPERIMENTAL STUDIES
a) RANDOMISED CONTROL TRIAL
Experimental studies can also be called intervention studies. They involve the active allocation to an exposure
or intervention group by the investigator. The group without the exposure or intervention acts as a control
group which may be given a placebo or another treatment. The occurrence of the outcome is then compared
between the intervention and control group as shown in the example below.
The best example of experimental studies is the randomised controlled trial where participants are randomly
allocated to either the exposure or control group. These provide the best evidence of a true causal association
5
between an exposure and an outcome, because when designed well (ensuring that the allocation of the
exposure is determined by chance alone) they minimise the possibility of selection bias and confounding.
Because of this, randomised controlled trials are considered the gold standard experimental study design.
Selection bias-Selection bias refers to error due to systematic differences in characteristics between those
who take part in a study and those who do not, such that:
the people who are selected to participate in a study are not representative of the reference
population or, in analytic studies,
the comparison groups are not comparable
Confounding is about alternative explanations for an association between an exposure and an outcome. It
occurs when there is unequal distribution of a risk factor for a disease/outcome between those exposed and
unexposed to the exposure of interest.
Advantages of RCT
Good randomization will “wash out” any population bias
Easier to blind/mask than observational studies
Results can be analyzed with well known statistical tools
Populations of participating individuals are clearly identified
Disadvantages of RCT
Expensive in terms of time and money
Volunteer biases: the population that participates may not be representative of the whole
Does not reveal causation
Loss to follow-up attributed to treatment
Fictitious Example
6
To determine how a new type of short wave UVA-blocking sunscreen affects the general health of skin in
comparison to a regular long wave UVA-blocking sunscreen, 40 trial participants were randomly separated into
equal groups of 20: an experimental group and a control group. All participants' skin health was then initially
evaluated. The experimental group wore the short wave UVA-blocking sunscreen daily, and the control group
wore the long wave UVA-blocking sunscreen daily.
After one year, the general health of the skin was measured in both groups and statistically analyzed. In the
control group, wearing long wave UVA-blocking sunscreen daily led to improvements in general skin health for
60% of the participants. In the experimental group, wearing short wave UVA-blocking sunscreen daily led to
improvements in general skin health for 75% of the participants.
2. OBSERVATIONAL STUDIES
In these types of studies, the investigator looks at the distribution or determinants of an outcome without
attempting to change the factors that may influence them. In other words nature is allowed to take its course
without any manipulation by the investigator. These studies can be 'individual-based' or 'population-based'.
a) CROSS-SECTIONAL STUDY
This is the simplest type of study design. Here the investigator measures the frequency of a particular
exposure(s) and/or outcome(s) in a defined population at one point in time. Cross-sectional studies can either
be descriptive or analytical.
A study design where one or more samples (called cohorts) are followed prospectively and subsequent status
evaluations with respect to a disease or outcome are conducted to determine which initial participants
exposure characteristics (risk factors) are associated with it. As the study is conducted, outcome from
participants in each cohort is measured and relationships with specific characteristics determined
Subjects in cohorts can be matched, which limits the influence of confounding variables
Standardization of criteria/outcome is possible
Easier and cheaper than a randomized controlled trial (RCT)
7
Design pitfalls to look out for in cohort ⁄ longitudinal study
The cohorts need to be chosen from separate, but similar, populations.
How many differences are there between the control cohort and the experiment cohort?
Will those differences cloud the study outcomes?
C) CASE-CONTROL STUDY
In this type of study, a group of individuals who already have the outcome of interest (the cases) are identified
from a specific population. A control group of individuals without the outcome is then selected from the same
population from which the cases arose. The prevalence of the past exposure is then compared between cases
and controls.
If the level of exposure is higher in the cases than in controls this may indicate that the exposure is a
risk factor for the disease.
If the level of exposure is lower in cases than controls, this may indicate that the exposure is
protective against the disease.
8
Example of a case-control study
A study looking at potential risk factors for obesity recruited all new diagnosed cases of obesity at a local clinic.
One control for each case was recruited among other people attending the same clinic but without obesity.
Questionnaires were administered to gather information on life style, demographic and socio-economic -
related exposures and levels of exposure compared between cases and controls.
Retrospective studies have more problems with data quality because they rely on memory and people
with a condition will be more motivated to recall risk factors (also called recall bias).
Not good for evaluating diagnostic tests because it’s already clear that the cases have the condition
and the controls do not
It can be difficult to find a suitable control group
Care should be taken to avoid confounding, which arises when an exposure and an outcome are both strongly
associated with a third variable. Controls should be subjects who might have been cases in the study but are
selected independent of the exposure. Cases and controls should also not be "over-matched."
Is the control group appropriate for the population? Does the study use matching or pairing appropriately to
avoid the effects of a confounding variable? Does it use appropriate inclusion and exclusion criteria?
There is a suspicion that zinc oxide, the white non-absorbent sunscreen traditionally worn by lifeguards is
more effective at preventing sunburns that lead to skin cancer than absorbent sunscreen lotions. A case-
control study was conducted to investigate if exposure to zinc oxide is a more effective skin cancer prevention
measure. The study involved comparing a group of former lifeguards that had developed cancer on their
cheeks and noses (cases) to a group of lifeguards without this type of cancer (controls) and assess their prior
exposure to zinc oxide or absorbent sunscreen lotions.
This study would be retrospective in that the former lifeguards would be asked to recall which type of
sunscreen they used on their face and approximately how often. This could be either a matched or unmatched
study, but efforts would need to be made to ensure that the former lifeguards are of the same average age ,
and lifeguarded for a similar number of seasons and amount of time per season.
D) ECOLOGICAL STUDIES
Individual-based observational studies are concerned with data derived from individuals within a population.
Sometimes we may want to look more broadly at the effect of an exposure or an outcome at a population
level.
Ecological studies compare the level of exposure and outcome between groups rather than individuals. They
relate the rate of disease to an exposure that is assumed to apply to all individuals in that population group.
9
where the exposure (low levels of iodine in the soil) was more frequent. The prevalence of iodine deficiency
was found higher in populations with higher levels of iodine depletion in the soil.
10
TOPIC 2b: ETHICAL ISSUES IN HUMAN RESEARCH
Objectives:
• Understand ethical principles for research involving humans
• Explore and discuss the ethical issues in examples of research
Definition of ethics:
Difference between right and wrong
Ethical principles regarding human experimentation developed for the medical community
1. Nuremberg code
2. Declaration of Helsinki
Treatment Withheld
There were no proven treatments for syphilis when the study began. When penicillin became the standard
treatment for the disease in 1947 the medicine was withheld as a part of the treatment for both the
experimental group and control group.
IRB Membership
IRB members should be collectively qualified to review the scientific, medical and ethical aspects of the trial.
12
TOPIC 3a: METHODS OF DATA COLLECTION, ORGANIZATION AND
PRESENTATION
Objectives
1. Identify the different methods of data organization and presentation
2. Understand the criterion for the selection of a method to organize and present data
3. Identify the different methods of data collection and criterion that we use to select a method of data
collection
4. Define a questionnaire, identify the different parts of a questionnaire and indicate the procedures to
prepare a questionnaire
Introduction
Before any statistical work can be done data must be collected. Depending on the type of variable and the
objective of the study different data collection methods can be employed.
1. Observation – Observation is a technique that involves systematically selecting, watching and recoding
behaviors of people or other phenomena and aspects of the setting in which they occur, for the purpose of
getting (gaining) specified information. It includes all methods from simple visual observations to the use of
high level machines and measurements, sophisticated equipment or facilities, such as radiographic,
biochemical, X-ray machines, microscope, clinical examinations, and microbiological examinations.
Outline the guidelines for the observations prior to actual data collection.
Disadvantages: Investigators or observer’s own biases, prejudice, desires, and etc. and needs more resources
and skilled human power during the use of high level machines.
• What exactly do we want to know, according to the objectives and variables we identified earlier?
Is questioning the right technique to obtain all answers, or do we need additional techniques, such as
observations or analysis of records?
Of whom will we ask questions and what techniques will we use?
Do we understand the topic sufficiently to design a questionnaire, or do we need some loosely structured
interviews with key informants or a focus group discussion first to orient ourselves?
• Are our informants mainly literate or illiterate? If illiterate, the use of self-administered questionnaires is not
an option.
13
CCCCCCCCCCCCCCCCCCCCCCCCCCC4• How large is the sample that will be interviewed? Studies with many
respondents often use shorter, highly structured questionnaires, whereas smaller studies allow more flexibility
and may use questionnaires with a number of open-ended questions.
Once the decision has been made Interviews may be less or more structured. Unstructured interview is
flexible, the content wording and order of the questions vary from interview to interview.
The investigators only have idea of what they want to learn but do not decide in advance exactly what
questions will be asked, or in what order.
For structured interviews, a more standardized technique may be used, the wording and order of the
questions being decided in advance. This may take the form of a highly structured interview, in which the
questions are asked orderly, or a self administered questionnaire, in which case the respondent reads the
questions and fill in the answers by himself (sometimes in the presence of an interviewer who ‘stands by’ to
give assistance if necessary).
Standardized methods of asking questions are usually preferred in public health research, since they provide
more assurance that the data will be reproducible. Less structured interviews may be useful in a preliminary
survey, where the purpose is to obtain information to help in the subsequent planning of a study rather than
factors for analysis, and in intensive studies of perceptions, attitudes, motivation and affective reactions.
Unstructured interviews are characteristic of qualitative (non-quantitative) research. The use of self-
administered questionnaires is simpler and cheaper; such questionnaires can be administered to many persons
simultaneously, and unlike interviews, can be sent by post. On the other hand, they demand a certain level of
education and skill on the part of the respondents; people of a low socio-economic status are less likely to
respond to a mailed
questionnaire.
In interviewing using questionnaire, the investigator appoints agents known as enumerators, who go to the
respondents personally with the questionnaire, ask them the questions given there in, and record their replies.
They can be either face-to-face or telephone interviews.
Face-to-face and telephone interviews have many advantages. A good interviewer can stimulate and maintain
the respondent’s interest, and can create a rapport (understanding, concord) and atmosphere conducive to
the answering of questions. If anxiety aroused, the interviewer can allay it. If a question is not understood an
interviewer can repeat it and if necessary (and in accordance with
guidelines decided in advance) provide an explanation or alternative wording. Optional follow-up or probing
questions that are to be asked only if prior responses are inconclusive or inconsistent cannot easily be built
into self-administered questionnaires. In face-to-face interviews, observations can be made as well.
In general, apart from their expenses, interviews are preferable to self-administered questionnaire, with the
important proviso that they are conducted by skilled interviewers.
3. Mailed Questionnaire Method: Under this method, the investigator prepares a questionnaire containing a
number of questions pertaining the field of inquiry. The questionnaires are sent by post to the informants
together with a polite covering letter explaining the detail, the aims and objectives of collecting the
information, and requesting the respondents to cooperate by furnishing the correct replies and returning the
questionnaire duly filled in. In order to ensure quick response, the return postage expenses are usually borne
by the investigator.
The main problems with postal questionnaire are that response rates tend to be relatively low, and that there
may be under representation of less literate subjects.
4. Use of documentary sources: Clinical and other personal records, death certificates, published mortality
statistics, census publications, etc.
Examples include:
1. Official publications of Central Statistical Authority
2. Publication of Ministry of Health and Other Ministries
3. News Papers and Journals.
4. International Publications like Publications by WHO, World Bank, UNICEF
14
5. Records of hospitals or any Health Institutions.
During the use of data from documents, though they are less time consuming and relatively have low cost,
care should be taken on the quality and completeness of the data. There could be differences in objectives
between the primary author of the data and the user.
A focus group is a small group of six to ten people led through an open discussion by a skilled moderator.
The group needs to be large enough to generate rich discussion but not so large that some participants are
left out.
The focus group moderator nurtures disclosure in an open and spontaneous format. The moderator’s goal
is to generate a maximum number of different ideas and opinions from as many different people in the
time allotted.
The ideal amount of time to set aside for a focus group is anywhere from 45 to 90 minutes. Beyond that
most groups are not productive and it becomes an imposition on participant time.
Focus groups are structured around a set of carefully predetermined questions – usually no more than 10
– but the discussion is free-flowing. Ideally, participant comments will stimulate and influence the thinking
and sharing of others. Some people even find themselves changing their thoughts and opinions during the
group.
A homogeneous group of strangers comprise the focus group. Homogeneity levels the playing field and
reduces inhibitions among people who will probably never see each other again.
It takes more than one focus group on any one topic to produce valid results – usually three or four. You’ll
know you’ve conducted enough groups (with the same set of questions) when you’re not hearing anything
new anymore, i.e. you’ve reached a point of saturation.
Focus group participants won’t have a chance to see the questions they are being asked. So, to make sure they
understand and can fully respond to the questions posed, questions should be:
Short and to the point
Focused on one dimension each
Unambiguously worded
Open-ended or sentence completion types
Non-threatening or embarrassing
Worded in a way that they cannot be answered with a simple “yes” or “no” answer (use “why” and
“how” instead)
15
1. Engagement questions: introduce participants to and make them comfortable with the topic of discussion
2. Exploration questions: get to the meat of the discussion
3. Exit question: check to see if anything was missed in the discussion
Engagement questions:
1.What is your favourite food?
2. What do you notice when you look at other people’s food?
Exploration Questions:
1. Who in particular has influenced your food choices?
2. What are the pros and cons of eating ''bad food''?
3. When you eat ''bad food'', how do you feel?
4. How do you feel when told about possible dangers of eating ''bad food''?
Exit question:
1. Is there anything else you would like to say about why you do or do not eat ''bad food''
The statistical data may be classified under two categories, depending upon the sources.
1) Primary data 2) Secondary data
Primary Data: are those data, which are collected by the investigator himself for the purpose of a specific
inquiry or study. Such data are original in character and are mostly generated by surveys conducted by
individuals or research institutions.
The first hand information obtained by the investigator is more reliable and accurate since the investigator can
extract the correct information by removing doubts, if any, in the minds of the respondents regarding certain
questions. High response rates might be obtained since the answers to various questions are obtained on the
spot. It permits explanation of questions concerning difficult subject matter.
Secondary Data: When an investigator uses data, which have already been collected by others, such data are
called "Secondary Data". Such data are primary data for the agency that collected them,
and become secondary for someone else who uses these data for his own purposes.
The secondary data can be obtained from journals, reports, government publications, publications of
professionals and research organizations.
Secondary data are less expensive to collect both in money and time. These data can also be better utilized
and sometimes the quality of such data may be better because these might have been collected by persons
who were specially trained for that purpose.
On the other hand, such data must be used with great care, because such data may also be full of errors due
to the fact that the purpose of the collection of the data by the primary agency may have been different from
the purpose of the user of these secondary data.
Secondly, there may have been bias introduced, the size of the sample may have been inadequate, or there
may have been arithmetic or definition errors, hence, it is necessary to critically investigate the validity of the
secondary data.
16
In general, the choice of methods of data collection is largely based on the accuracy of the information they
yield. In this context, ‘accuracy’ refers not only to correspondence between the information and objective
reality - although this certainly enters into the concept - but also to the information’s relevance.
This issue is the extent to which the method will provide a precise measure of the variable the investigator
wishes to study.
The selection of the method of data collection is also based on practical considerations, such as:
1) The need for personnel, skills, equipment, etc. in relation to what is available and the urgency with which
results are needed.
2) The acceptability of the procedures to the subjects - the absence of inconvenience, unpleasantness, or
untoward consequences.
3) The probability that the method will provide a good coverage, i.e. will supply the required information
about all or almost all members of the population or sample. If many people will not
know the answer to the question, the question is not an appropriate one.
The investigator’s familiarity with a study procedure may be a valid consideration. It comes as no particular
surprise to discover that a scientist formulates problems in a way which requires for their
solution just those techniques in which he himself is specially skilled
Open-ended questions
Open-ended questions permit free responses that should be recorded in the respondent’s own words. The
respondent is not given any possible answers to choose from.
Closed Questions
Closed questions offer a list of possible options or answers from which the respondents must choose. When
designing closed questions one should try to:
Offer a list of options that are exhaustive and mutually exclusive
Keep the number of options as few as possible.
Have you ever gone to the local village health worker for treatment?
1. Yes
2. No
Closed questions may also be used if one is only interested in certain aspects of an issue and does not want to
waste the time of the respondent and interviewer by obtaining more information than one needs.
Requirements of questions
17
Must have face validity –the question that we design should be one that give an obviously valid and
relevant measurement for the variable. For example, it may be self-evident that records kept in an
obstetrics ward will provide a more valid indication of birth weights than information obtained by
questioning others.
Must be clear and unambiguous – the way in which questions are worded can ‘make or break’ a
questionnaire. They must be phrased in language that it is believed the respondent will understand, and
that all respondents will understand in the same way.
To ensure clarity, each question should contain only one idea; ‘double-barrelled’ questions like ‘Do you
take your child to a doctor when he has a cold or has diarrhoea?’ are difficult to answer, and the answers
are difficult to interpret.
Must not be offensive – whenever possible it is wise to avoid questions that may offend the respondent,
for example those that deal with intimate matters, those which may seem to expose the respondent’s
ignorance, and those requiring him to give a socially unacceptable answer.
The questions should be fair - They should not be phrased in a way that suggests a specific answer, and
should not be loaded. Short questions are generally regarded as preferable to long ones. Sensitive
questions - It may not be possible to avoid asking ‘sensitive’questions that may offend respondents, e.g.
those that seem to expose the respondent’s ignorance. In such situations the interviewer should do it very
carefully and wisely
Usually the questionnaire will need some further adaptation before we can use it for actual data
collection.
Step1: CONTENT
Take your objectives and variables as your starting point.
Decide what questions will be needed to measure or to define your variables and reach your objectives. When
developing the questionnaire, you should reconsider the variables you have chosen, and, if necessary, add,
drop or change some. You may even change some of your objectives at this stage.
18
Check whether each question measures one thing at a time.
For example, the question, ''How large an interval would you and your husband prefer between two
successive births?'' would better be divided into two questions because husband and wife may have
different opinions on the preferred interval.
19
METHODS OF DATA ORGANIZATION AND
PRESENTATION
The data collected in a survey is called raw data. In
most cases, useful information is not immediately
evident from the mass of unsorted data. Collected data
need to be organized in such a way as to condense the
information they contain in a way that will show
patterns of variation clearly. Precise methods of
analysis can be decided up on only when the
characteristics of the data are understood.
A) Frequency Distributions
For data to be more easily appreciated and to draw
quick comparisons, it is often useful to arrange the
data in the form of a table, or in one of a number of
different graphical forms. When analysing voluminous
data collected from say, a health centres' records, it is
quite useful to put them into compact tables. Quite
often, the presentation of data in a meaningful way is
done by preparing a frequency distribution. If this is
not done the raw data will not present any meaning
and any pattern in them (if any) may not be detected.
20
Simple frequency distribution
A frequency distribution simply tells how often a
variable takes on each of its possible values. For
quantitative variables with many possible values the
possible values are typically binned or grouped into
intervals. Let’s take an example of the number of males
and females in this class:
Example: The distribution of Gender in HMN 1 2021
class in Makerere University
Gender Number of Frequency (%)
participants
Males 72 40
Females 106 60
Total 178
Construction of tables
Although there are no hard and fast rules to follow, the
following general principles should be addressed in
constructing tables.
1. Tables should be as simple as possible.
2. Tables should be self-explanatory. I.e.
• Title should be clear and to the point (a good title
answers: what? when? where? how classified) and it
should be placed above the table.
• Each row and column should be labelled.
• Numerical entities of zero should be explicitly written
rather than indicated by a dash. Dashed are reserved
for missing or unobserved data.
• Totals should be shown either in the top row and the
first column or in the last row and last column.
3. If data are not original, their source should be given
in a footnote.
B) Construction of graphs
The choice of the particular form among the different
possibilities will depend on personal choices and/or the
type of the data.
• Bar charts and pie chart are commonly used for
qualitative or quantitative discrete data.
22
• Histograms, frequency polygons are used for
quantitative continuous data.
Bar Chart
Bar diagrams are used to represent and compare the
frequency distribution of discrete variables and
attributes or categorical series. When we represent
data using bar diagram, all the bars must have
equal width and the distance between bars must be
equal. There are different types of bar diagrams, the
most important ones are:
23
-Simple bar chart: It is a one-dimensional diagram in
which the bar represents the whole of the magnitude.
-Pie chart
25
Topic 4: Summarizing data
Objectives
1. Identify the different methods of data summarization
2. Compute appropriate summary values for a set of data
3. Appreciate the properties and limitations of summary values
Introduction
The best way to work with data is to summarize and organize them. Numbers that have not been summarized
and organized are called raw data.
Descriptive measures
A descriptive measure is a single number that is used to describe a set of data. Descriptive measures include
A) measures of central tendency/location and
B) measures of dispersion.
26
Properties of the mean
a. Uniqueness
For a given set of data there is one and only
one mean.
b. Simplicity
The mean is easy to calculate.
c. Affected by extreme values
the mean is influenced by each value.
Therefore, extreme values can distort the mean. This
means that it is sensitive to extreme values which at
times may not be typical of the data set as a whole.
For example:
For example, if the first infant were in the above data
happened to be a premature infant weighing 1.5kg
rather than 3.2kg, then the arithmetic mean of the
sample would be reduced. The arithmetic mean is a
poor measure of central location in these types of
sample, since it does not reflect the centre of sample.
Nevertheless, the arithmetic mean is by far the most
widely used measure of central-location.
2. Median
The median is the value that divides the set of data
into two equal parts. The number of values equal to or
greater than the median equals the number of values
less than or equal to the median.
i.e For example, if the first infant were in the above data happened to be a premature infant weighing
1.5kg rather than 3.2kg, then the median of the sample would not change.
Mode
The mode is the value that occurs most often in a set of data. It is possible to have more than one mode or no
mode.
As a general rule:
For symmetric data, the mean and the median are
the same.
With skewed data, the mean can be heavily
influenced by the random presence of
a/some extreme observation(s).
28
• In order to still get a good idea about the location of
the data, one then prefers
the use of the median over the mean:
Skewness: If extremely low or extremely high observations are present in a distribution, then the mean tends
to shift towards those scores.
b) Positively skewed distribution: Occurs when the majority of scores are at the left end of the curve and a
few extreme large scores are scattered at the right end.
c) Symmetrical distribution: It is neither positively nor negatively skewed. A curve is symmetrical if one half of
the curve is the mirror image of the other half.
NB: Numerical variables, such as the distribution body weights, heights, blood pressure measurements in the
general population tend to approximately follow a bell-shaped curve, commonly referred to as the Normal
distribution or Gaussian distribution. This distribution is symmetrical about its mean.
While the mean, median, etc. give useful information about the center of the data, we also need to know how
“spread out” the numbers are about the center.
The two data sets given above have a mean of 50, but obviously set 1 is more “spread out” than set 2. The
question is how do we express this numerically?
The object of measuring this scatter or dispersion is to obtain a single summary figure which adequately
exhibits whether the distribution is compact or spread out.
Some of the commonly used measures of dispersion (variation) are: Range, variance, standard deviation and
coefficient of variation.
29
1. Range
The range is defined as the difference between the
highest and smallest observation in the data. It is the
crudest measure of dispersion. The range is a measure
of absolute dispersion and as such cannot be usefully
employed for comparing the variability of two
distributions expressed in different units.
Where :
Xmax = highest (maximum) value in the given
distribution.
Xmin = lowest (minimum) value in the given
distribution.
where:
∑=Sum of
X=individual values in the data set
µ=mean of all values in the data set
N=Number of values in the data set
31
First, calculate the difference of each data point from
the mean, and square the result of each:
(2-4)2=(-2) 2=4
(3-4)2=(-1) 2=1
(5-4)2=(1) 2=1
(6-4)2=(2) 2=4
Next, calculate the mean of these values, and take the
square root:
√4+1+1+4 =1.58
4
This quantity is the population standard deviation, and
is equal to the square root of the variance
Variance:
The square of the standard deviation is called the variance. The variance is a very useful measure of variability
because it uses the information provided by every observation in the sample and also it is
very easy to handle mathematically.
Definition: The coefficient of variation (CV) is defined by: 100%* standard deviation
mean
The coefficient of variation is most useful in comparing the variability of several different samples,
each with different means.
This is because a higher variability is usually expected when the mean increases, and the CV is a
measure that accounts for this variability.
Example: Compute the CV for the birth weight data shown below.
32
Topic 6: Sampling methods
OBJECTIVES
1. Define population and sample and understand the different sampling terminologies
2. Differentiate between probability and Non-Probability sampling methods and apply different
techniques of sampling
3. Understand the importance of a representative sample
4. Differentiate between random error and bias
5. Enumerate advantages and limitations of the different sampling methods
INTRODUCTION
Sampling involves the selection of a number of a study units from a defined population. The
population is too large for us to consider collecting information from all its members. If the whole
population is taken there is no need of statistical inference. Usually, a representative subgroup of
the population (sample) is included in the investigation. A representative sample has all the
important characteristics of the population from which it is drawn.
Advantages of samples
• cost - sampling saves time, labour and money
• quality of data - more time and effort can be spent on getting reliable data on each individual
included in the sample.
Example
- if the objective is to determine the availability of latrines, then the study unit would be the
household
-if the objective is to determine the prevalence of malnutrition, then the study unit would be the
individual child.
Sampling frame - the list of all the units in the reference population, from which a sample is to be
picked. e.g list of all households
Sampling interval) - the ratio of the number of units in the sample to the number of units in the
reference population
(n/N)
33
SAMPLING METHODS (TWO BROAD DIVISIONS)
1. NON-PROBABILITY SAMPLING METHODS
- Used when a sampling frame does not exist
- No random selection (unrepresentative of the given population)
- Inappropriate if the aim is to measure variables and generalize findings obtained from a sample to
the population.
b) Quota sampling: is a method that ensures that a certain number of sample units from different
categories with specific characteristics are represented. In this method the investigator interviews as
many people in each category of study unit as he can find until he has filled his quota.
Both the above methods do not claim to be representative of the entire population.
However, it is costly to conduct SRS. Moreover, minority subgroups of interest in the population may
not be present in the sample in sufficient numbers for study.
"Lottery” method: for a small population it may be possible to use the “lottery” method: each unit in
the population is represented by a slip of paper, these are put in a box and mixed, and a sample of
the required size is drawn from the box.
Table of random numbers: if there are many units, however, the above technique soon becomes
laborious. Selection of the units is greatly facilitated and made more accurate by using a set of
random numbers in which a large number of digits is set out in random order.
The property of a table of random numbers is that, whichever way it is read, vertically in columns or
horizontally in rows, the order of the digits is random. Nowadays, any scientific calculator has the
same facilities.
34
B) SYSTEMATIC SAMPLING
Individuals are chosen at regular intervals ( for example, every k th) from the sampling frame. The first
unit to be selected is taken at random from among the first k units. For example, a systematic
sample is to be selected from 1200 students of a school. The sample size is decided to be 100. Step
1: calculate the sampling interval: The sampling fraction or interval is: 100 /1200 = 1/12. Hence, the
sample interval is 12.
The number of the first student to be included in the sample is chosen randomly, for example by
blindly picking one out of twelve pieces of paper, numbered 1 to 12. If number 6 is picked, every
twelfth student will be included in the sample, starting with student number 6, until 100 students
are selected. The numbers selected would be 6,18,30,42,etc.
Advantages
• Systematic sampling is usually less time consuming and easier to perform than simple random
sampling.
• Unlike SRS, systematic sampling can be conducted without a sampling frame (useful in some
situations where a sampling frame is not readily available).
Eg., In patients attending a health center, where it is not possible to predict in advance who will be
attending.
Disadvantages
• If there is any sort of cyclic pattern in the ordering of the subjects which coincides with the
sampling interval, the sample will not be representative of the population.
C) STRATIFIED SAMPLING
It is appropriate when the distribution of the characteristic to be studied is strongly affected by
certain variable (diverse population). The population is first divided into groups (strata)
according to a characteristic of interest (eg., sex, geographic area, prevalence of disease, etc.). A
separate sample is then taken independently from each stratum, by simple random or systematic
sampling.
• Proportional allocation - if the same sampling fraction is used for each stratum.
• Non- proportional allocation - if a different sampling fraction is used for each stratum or if the
strata are unequal in size and a fixed number of units is selected from each stratum.
Advantages
- The representativeness of the sample is improved. That is, adequate representation of minority
subgroups of interest can be ensured by stratification and by varying the sampling fraction between
strata as required.
Disadvantages
- Sampling frame for the entire population has to be prepared separately for each stratum.
D) CLUSTER SAMPLING
In this sampling scheme, selection of the required sample is done on groups of study units (clusters)
instead of each study unit individually.
35
The sampling unit is a cluster, and the sampling frame is a list of these clusters.
Procedure
- The reference population (homogeneous) is divided into clusters.
These clusters are often geographic units (eg districts, villages, etc.)
- A sample of such clusters is selected
- All the units in the selected clusters are studied
It is preferable to select a large number of small clusters rather than a small number of large
clusters.
Advantages
A list of all the individual study units in the reference population is not required. It is sufficient to
have a list of clusters.
Disadvantages
It is based on the assumption that the characteristic to be studied is uniformly distributed
throughout the reference population, which may not always be the case. Hence, sampling error is
usually higher than for a simple random sample of the same size.
E) MULTI-STAGE SAMPLING
This method is appropriate when the reference population is large and widely scattered . Selection is
done in stages until the final sampling unit (eg., households or persons) are arrived at. The primary
sampling unit (PSU) is the sampling unit (usually large size) in the first sampling stage. The secondary
sampling unit (SSU) is the sampling unit in the second sampling stage, etc.
Example - The PSUs could be villages and the SSUs could be households.
Advantages
- Cuts the cost of preparing sampling frame
Disadvantages
- Sampling error is increased compared with a simple random sample.
Multistage sampling gives less precise estimates than simple random sampling for the same sample
size, but the reduction in cost usually far outweighs this, and allows for a larger sample size.
ERRORS IN SAMPLING
When we take a sample, our results will not exactly equal the correct results for the whole
population. That is, our results will be subject to errors.
36
eliminate or reduce the non-sampling error (bias) by careful design of the sampling procedure and
not by increasing the sample size.
Example: If you take male students only from the year 1 Human Nutrition class in order to
determine the proportion of those who like cooking, you would result in an underestimate, since
males are less likely to cook. Increasing the number of male students would not remove the bias.
37
Topic 8: Hypothesis Testing
Objectives
1. Understand the concepts of null and alternative hypothesis
2. Explain the meaning and application of statistical significance
3. Differentiate between type I and type II errors
4. Explain the meaning and application of P – values
5.Interpret and explain a p-value
6. Understand the application of confidence intervals and be able to calculate them
Introduction
Hypothesis testing helps us to measure the strength of the evidence which the data supply
concerning some proposition of interest.
Definition :A statistical hypothesis is an assumption or a statement which may or may not be true
concerning one or more populations.
Examples of hypothesis
1) The mean height of the girls in the BSC Human Nutrition Class is 1.53m.
2) There is no difference between the mean weights of boys and girls in the BSC Human Nutrition
Class
Simple
A simple hypothesis contains one predictor and one outcome variable, e.g. eating mandazi
increases the risk of obesity in P.7 children..
Here the single predictor variable is mandazi and the outcome variable is obesity.
A complex hypothesis contains more than one predictor variable or more than one outcome
variable, e.g., eating mandazi and studying in urban schools are associated with an increased
risk of obesity .
Here there are 2 predictor variables, i.e., mandazi and studying in urban schools, and one
outcome variable, i.e., obesity. Complex hypothesis like this cannot be easily tested with a
single statistical test and should always be separated into 2 or more simple hypotheses.
Specific
A specific hypothesis leaves no ambiguity about the subjects and variables, or about how the
test of statistical significance will be applied. It uses concise operational definitions that
summarize the nature and source of the subjects and the approach to measuring variables .
38
TYPES OF HYPOTHESES
For the purpose of testing statistical significance, hypotheses are classified by the way they describe
the expected differences or relationships between the study groups or variables.
The proposition that there is an difference between P.7 children who eat mandazi are a
obese as those who do not eat mandazi is called the alternative hypothesis. The alternative
hypothesis cannot be tested directly; it is accepted by exclusion if the test of statistical
significance rejects the null hypothesis.
HYPOTHESIS TESTING
Hypothesis testing is the process of using samples to make inferences about a population. Usually it
is not possible to study the entire population. In most sample surveys we do not know the true
population measure, we can only infer from our "sample statistic", also referred as a point estimate.
39
• A cut-off of P= <.05 means that in about 5 out of 100 (1 in 20) experiments, a result
would appear significant just by chance i.e that there is less than one in 20 chance
that the difference seen is could have arisen by chance if there was no true
difference
• We can use other P values for example 0.01
• We reject the null hypothesis when the P-value is less than or equal to 0.05 i.e
(p<0.05), implying that the two means (Mean1 and Mean2) are significantly
different.
• On the contrary, when the P-value is greater than 0.05 (P > 0.05) then we cannot
reject the null hypothesis.
d. Choose a random sample from the appropriate population and compute appropriate sample
statistics/sample evidence (e.g mean) to decide whether to whether the established null
hypothesis is rejected or not rejected.
e. Make a conclusion; for example:
• The difference between mean 1 and mean 2 is not equal to zero; therefore the null
hypothesis that the null hypothesis is rejected
• There is no a difference between mean 1 and mean 2; therefore the null hypothesis
is rejected
• The mean weight of P.7 children who eat mandazi is higher than for those who do
not eat mandazi; therefore the null hypothesis that there is no difference between
p.7 children who eat mandazi and those who do not is rejected.
In some ways, the investigator’s problem is similar to that faced by a judge judging a
defendant.
The absolute truth whether the defendant committed the crime cannot be determined.
Instead, the judge begins by presuming innocence — the defendant did not commit the
crime: in our case we begin by saying there is no difference......
The judge must decide whether there is sufficient evidence to reject the presumed
innocence of the defendant;
A judge can err, however, by convicting a defendant who is innocent, or by failing to convict
one who is actually guilty.
In similar fashion, the investigator starts by presuming the null hypothesis, or no
association/difference between the predictor and outcome variables in the population.
Based on the data collected in his sample, the investigator uses statistical tests to determine
whether there is sufficient evidence to reject the null hypothesis in favour of the alternative
hypothesis that there is an association in the population.
The judge may however convict the defendant when he is not guilty or may fail to convict
him when he is guilty
similarly, the investigator may reject the null hypothesis when it is correct in the population
or fail to reject it when it is false in the population
A Type I error is made when HO (null hypothesis) is true but rejected. That is if there is no
association but the researcher concludes that there is an association between the
independent and dependent variables.
A Type II error is made when HO is false but we fail to reject it, in other words, when we say
that there is no relationship but when actually a relationship exists.
40
Just like a judge’s conclusion, an investigator’s conclusion may be wrong. Sometimes, by
chance alone or a sample is not representative of the population.
Thus the results in the sample do not reflect reality in the population, and the random error
leads to an erroneous inference.
A type I error (false-positive) occurs if an investigator rejects a null hypothesis that is actually
true in the population;
A type II error (false-negative) occurs if the investigator fails to reject a null hypothesis that is
actually false in the population.
Although type I and type II errors can never be avoided entirely, the investigator can reduce
their likelihood by increasing the sample size (the larger the sample, the lesser is the
likelihood that it will differ substantially from the population).
o The likelihood that a study will be able to detect an association between a predictor
variable and an outcome variable depends, of course, on the actual magnitude of
that association in the target population. If it is large it will be easy to detect in the
sample.
o Conversely, if the size of the association is small it will be difficult to detect in the
sample. Unfortunately, the investigator often does not know the actual magnitude
of the association — one of the purposes of the study is to estimate it. Instead, the
investigator must choose the size of the association that he would like to be able to
detect in the sample. This quantity is known as the effect size.
o Selecting an appropriate effect size is the most difficult aspect of sample size
planning. Sometimes, the investigator can use data from other studies or pilot tests
to make an informed guess about a reasonable effect size when there are no data
with which to estimates.
o The choice of the effect size is always somewhat arbitrary, and considerations of
feasibility are often paramount. When the number of available subjects is limited,
the investigator may have to work backward to determine whether the effect size
that his study will be able to detect with that number of subjects is reasonable.
Usually in research, the investigator establishes the maximum chance of making type I and
type II errors in advance of the study.
The probability of committing a type I error (rejecting the null hypothesis when it is actually
true) is called α (alpha) the other name for this is the level of statistical significance.
o If a study of mandazi eating and obesity is designed with α = 0.05, for example, then
the investigator has set 5% as the maximum chance of incorrectly rejecting the null
hypothesis (and erroneously inferring that eating mandazi is associated with
obesity).
o This is the level of reasonable doubt that the investigator is willing to accept when
he uses statistical tests to analyse the data after the study is completed.
o The probability of making a type II error (failing to reject the null hypothesis when it
is actually false) is called β (beta). The quantity (1 - β) is called power, the probability
of observing an effect in the sample (if one), of a specified effect size or greater
exists in the population.
o If β is set at 0.10, then the investigator has decided that he is willing to accept a 10%
chance of missing an association of a given effect size between mandazi and obesity.
o This represents a power of 0.90, i.e., a 90% chance of finding an association of that
size.
o Ideally alpha and beta errors would be set at zero, eliminating the possibility of
false-positive and false-negative results.
o In practice they are made as small as possible. Reducing them, however, usually
requires increasing the sample size.
41
o Sample size planning aims at choosing a sufficient number of subjects to keep alpha
and beta at acceptably low levels without making the study unnecessarily expensive
or difficult.
o Many studies set alpha at 0.05 and beta at 0.20 (a power of 0.80). These are
somewhat arbitrary values, and others are sometimes used; the conventional range
for alpha is between 0.01 and 0.10; and for beta, between 0.05 and 0.20.
o In general the investigator should choose a low value of alpha when the research
question makes it particularly important to avoid a type I (false-positive) error, and
he should choose a low value of beta when it is especially important to avoid a type
II error.
42
THE NORMAL DISTRIBUTION CURVE AND STANDARD
DEVIATION
The normal distribution curve also called Also
known as Gaussian distribution after the
mathematician Karl Friedrich Gauss is a very useful
curve in statistics because many attributes, when a
large of measurements are taken, are
approximately distributed in this pattern. Many
human characteristics, such as height, weight of a
large number of people, follow the normal
distribution
You may be wondering what is “normal” about the
normal distribution.
The name arose from the historical derivation of
this distribution as a model for the errors made in
astronomical observations and other scientific
observations.
In this model the “average/mean” represents the
true or normal value of the measurement and
deviations from this are errors.
Standard deviation indicates how much a set of
values is spread around the average/mean
In a normal distribution curve a range of
o 1SD above and below the mean (abbreviated
as + 1SD) includes 68.2% of the values
o 2SD above and below the mean (abbreviated
as + 2SD) includes 95.4% of the values
43
o 3SD above and below the mean (abbreviated
as + 3SD) includes 99.7% of the values
Example:
1. Assuming a group of patients had a normally
distributed weight. The mean weight of the patients is
80kg and the SD is 5kg.
44
b) Calculate how much 95.4% of the patients will weigh
in the sample.
We know that 95.4% of the patients represents
patients within +2SD of the sample
Example2 :
45
1SD below the sample =mean-1SD=80-(1*3) = 77
1SD above the sample =mean+1SD=80+ (1*3) = 83
That means 68.2% of the patients will weigh
between 77 and 83kg.
46
mean one goes (in either direction) the fewer the
scores;
5. Normal distributions are defined by two parameters,
the mean (μ) and the standard deviation (σ).
CONFIDENCE INTERVALS:
Confidence intervals are used when instead of simply wanting to know the mean value of a
sample, we want a range that is likely to contain the true population.
Example 1: For example we may be interested in increasing the mean weights of malnourished
children after feeding them a nutritious diet. From the sample of the children in the experiment,
we can work out the mean change in their weight. But this mean will only be for that particular
sample. If we took another group of patients we would not expect to get exactly the same value
because chance can also affect the change in weight.
Example 2: Suppose the mean weight among sixteen p.7 school children is 30 kg; this sample
mean is called a point estimate, and it is used to estimate the corresponding population mean.
If repeated samples are taken from the same population there is no reason to assume that the
population mean will be exactly equal to the sample mean.
A confidence interval is a range around a measurement that conveys how precise the
measurement is.
The commonest used in biostatics is a 95% C1.
With a 95 percent confidence interval, you have a 5 percent chance of being wrong.
With a 90 percent confidence interval, you have a 10 percent chance of being wrong.
The 95%CI gives a range around the mean where we expect the "true" (population) mean to lie,
using the formulae:
The above interval is called the 95% confidence interval (CI) for a population mean.
Note that the width of the CI depends on the sample size and on the variation of data values.
A very wide interval may indicate that more data should be collected before anything very
definite can be said about the population mean.
Example 1:
Suppose the mean weight of 16 children in P.7 is 30kg, with an SD of ± 5 kg. What is the 95% CI for
the population mean?
95% CI = sample mean ± 2SD/√n
=30± 2(5)/ √16
=30±10/4
=30±2.5 kg
95%CI=30 (27.5 to 32.5)
The above result tells us that we are 95% confident that the range 27.5 to 32.5 (i.e ± 2.5)
contains the true mean weight of the P.7 children.
We can also say that if we take another sample of P.7 children, 95 times out of 100 the true
mean weight will lie between 27.5 and 32.5 kg.
47
A CI tells you about how stable the estimate is. A stable estimate is one that would be close to
the same value if the survey were repeated.
An unstable estimate is one that would vary from one sample to another. Wider confidence
intervals in relation to the estimate itself indicate instability.
The size of the confidence interval is related to the sample size of the study: the larger the
sample size, the smaller the CI
NOTE:
Usually a P value and 95 % CI will give the same interpretation of whether there is a significant
difference or not. See Example below
Hypothesis 1: There is no difference in number of Patients satisfied with consultation between the
two Nutritionists (X,Y)
Ho: μ1= μ2
With a p-value of 0.38, we conclude that there is no sufficient evidence against the null hypothesis, since the
p-value is above 0.05. Therefore, we do not reject the null hypothesis and conclude that there is no difference
in satisfaction between the two nutritionists.
With respect to the CIs, 95% C1=90%-97%(X); 95% C1=82%-96% (Y); since there is an overlap between the two
confidence intervals, we do not reject the null hypothesis, both n nutritionists gave satisfactory performance.
NOTE:
Finding that an effect is significant does not tell you about how large or important the effect is.
Do not confuse statistical significance with practical significance. A small effect can be highly
significant if the sample size is large enough.
48
Finding that an effect is statistically significant signifies that the effect is real and not due to
chance
Getting a p value lower than the alpha level allows you to reject the null hypothesis, but
getting a p value greater than the alpha level does not prove that the null hypothesis is true.
1. Misconception: The probability value is the probability that the null hypothesis is false.
Proper interpretation: The probability value is the probability of a result as extreme or more
extreme given that the null hypothesis is true. It is the probability of the data given the null
hypothesis. It is not the probability that the null hypothesis is false.
Proper interpretation: A low probability value indicates that the sample outcome (or one more
extreme) would be very unlikely if the null hypothesis were true. A low probability value can
occur with small effect sizes, particularly if the sample size is large.
3. Misconception: A non-significant outcome means that the null hypothesis is probably true.
Proper interpretation: A non-significant outcome means that the data do not conclusively
demonstrate that the null hypothesis is false.
Type 11 error
In this type of error the researcher fails to reject a false null hypothesis.
Unlike a Type I error, a Type II error is not really an error. When a statistical test is not
significant, it means that the data do not provide strong evidence that the null hypothesis is
false.
Lack of significance does not support the conclusion that the null hypothesis is true. Therefore, a
researcher should not make the mistake of incorrectly concluding that the null hypothesis is true
49
when a statistical test was not significant (i.e observe nothing in our sample when it exists in the
population eg. smoking does not cause cancer). Instead, the researcher should consider the test
inconclusive.
Contrast this with a Type I error in which the researcher erroneously concludes that the null
hypothesis is false when, in fact, it is true, a Type II error can only occur if the null hypothesis is
false. If the null hypothesis is false, then the probability of a Type II error is called β (beta). The
probability of correctly rejecting a false null hypothesis equals 1- β and is called power.
The probability value is the proportion of times that you would get a difference in your sample
as large or larger than the one you found if the null hypothesis were actually true. Thus, lower
probability values make you more confident that the null hypothesis is false
50
Topic 9: Determining sample size
ATTRIBUTES OF A SAMPLE
Every individual in the chosen population should have an equal chance to be included in the
sample.
Ideally, choice of one participant should not affect the chance of another's selection (hence we
try to select the sample randomly – but it is important to note that random sampling does not
describe the sample or its size as much as it describes how the sample is chosen).
If we include very few subjects in a study, the results cannot be generalized to the population as
this sample will not represent the size of the target population. Further, the study then may not
be able to detect the difference between test groups.
On the other hand, if we study more subjects than required, we put more individuals to the risk
of the intervention, also making the study unethical, and waste precious resources, including the
researchers’ time.
The calculation of an adequate sample size thus becomes crucial in any study and is the process
by which we calculate the optimum number of participants required to be able to arrive at
ethically and scientifically valid results.
Some more factors that can be considered while calculating the final sample size include the
expected drop-out rate, an unequal allocation ratio, and the objective and design of the study.
A. LEVEL OF SIGNIFICANCE
• This is the p value that we set prior to starting a study. When we say, for example, we will accept
a p<0.05 as significant, we mean that we are ready to accept that the probability that the result
is observed due to chance (and NOT due to our intervention) is 5%.
• To put it in different words, we are willing to accept the detection of a difference 5 out of 100
times when actually no difference exists. or we are accepting that one in 20 times (that is 5%)
we will miss a real difference.
B.POWER
• Sometimes we may commit another type of error where we fail to detect a difference when
actually there is a difference (i.e observe nothing in our sample when it exists in the population
eg. smoking does not cause cancer).
• This is called the Type II error that detects a false negative difference, as against the one
mentioned above where we detect a false positive difference when no difference actually exists
51
or the Type I error (i.e observe something in our results but it does not exist in the population
e.g drinking water causes cancer).
• We must decide what is the false negative rate (Type II error) we are willing to accept to make
our study adequately powered to accept or reject our null hypothesis accurately.
• This false negative rate is referred to in statistics by the letter β. The “power” of the study then
is equal to (1 –β) and is the probability of failing to detect a difference when actually there is a
difference.
• If you increase the power of a study you decrease the chances of committing a Type II error.
• Usually most studies accept a power of 80%. This means that we are accepting that one in five
times (that is 20%) we will miss a real difference.
• Sometimes for large studies, the power is occasionally set at 90% to reduce to 10% the
possibility of a “false negative” result.
52
• Suppose we are studying the effect of an intervention on the weight and consider a population
with weights ranging from 45 to 100 kg.
• Naturally the standard deviation in this group will be great and we would need a larger sample
size to detect a difference between interventions, else the difference between the two groups
would be masked by the inherent difference between them because of the variance.
• If on the other hand, we were to take a sample from a population with weights between 80 and
100 kg we would naturally get a tighter and more homogenous group, thus reducing the
standard deviation and therefore the sample size.
The sample size calculated using the above formula is based on some conventions (Type I and II
errors) and few assumptions (effect size and standard variation).
The sample size ALWAYS has to be calculated before initiating a study and as far as possible
should not be changed during the study course.
The sample size calculation is also then influenced by a few practical issues, e.g., administrative
issues and costs.
Figure 3.1 New Cases of Illness from October 1, 2004–September 30, 2005
53
TOPIC 10: DESIGNING A RESEARCH PROPOSAL
Aims:
To outline the stages involved in creating a research question and a research proposal and define the
specific requirements involved in constructing each stage.
Learning outcomes:
By the end of this session you will be able to understand:
• The requirements for a good research question/hypothesis.
• The importance of having a clear question/hypothesis as a starting point for any study.
• The steps required to develop a research proposal.
Example:
Peanuts cause more food allergy in men than in women
Question: Do peanuts cause more food allergy in men than in women?
Developing a Proposal
In order to have the best chance of success requires:
• Forethought
• Preparation Time
• Critical Evaluation
• Review and compliance with the sponsor's requirements
1. Title
Give a brief title for your proposed study here. For example: An investigation into the causes of
malnutrition among children.
2. Abstract
Summary of the whole proposal
Scientific abstract versus lay summary
Lay statements: Use the langue of a 12 year old person reading the new vision
Scientific abstract:
Nano particles are small particles with one dimension less than 100 nm. These are used in more and
more modern applications and due to their small size but large surface area could have the potential
to be a hazard to human health and the lack of this knowledge has recently been highlighted by
several governmental agencies. Due to the small size of nano particles, many of these are airborne
and can thereby potentially be inhaled into the lower respiratory airways ....
Lay summary:
Nano particles are tiny particles and approximately 1,000 next to each other would be the same as
the width of a hair. These tiny nanoparticles are found in many modern products as pigments in
paint and cosmetics, anti-bacterial components in clothes and deodorants and in everyday traffic,
mainly in the exhaust from cars and lorries. Because of their tiny size these are often airborne and
able to reach into the deep areas of the airways ....
3. Introduction
• Literature review (selected)providing the background knowledge to understand the research
project
• Motivation for the study
• Problem and purpose
i.e Here, give a brief contextual background to your study. Set the stage for your study. Try to
capture your readers’ attention and focus on your study. Briefly tell us why this study should be
done. Use the proper referencing techniques! In the introduction, you should use evidence (facts,
figures, and works by other authors) to convince us that your chosen research topic is:
• Relevant
• Not already over-researched
• Feasible (in terms of scope, resources and a time frame)
• Ethically defensible.)
The introduction could be up to 1 page in length.
55
(If your study is predominantly quantitative, you will suggest hypotheses here. For example: role
confusion leads to care-giver burnout. Unrealistic expectations lead to care-giver burnout, etc. If
your study is qualitative or more exploratory and descriptive, you will formulate broad objectives.
For example: I want to understand the subjective experience of burnout amongst care-givers.) List a
few objectives.
Example
There is no difference in the risk of allergy from peanuts in men and women
8. Aims:
What are we going to investigate? The aim is to investigate the causes of malnutrition in
children.
9. Objectives:
How are we going to investigate it? These need to be
SMART I.e Specific, Measurable, Achievable, Realistic, Timely
Examples:
Examples:
1. To determine the nutritional status of children five years of age in katanga using
anthropometry
2. To determine the dietary diversity of children below two years of age in Katanga using the
24-hour recall method
Give us as much detail as possible here – for example I will administer a standard questionnaire to
the entire sample of caregivers for a month at their homes in Katanga. I will conduct personal
interviews with a sub-sample of caregivers, using a semi-formal interview schedule as attached in
appendix A. I will make appointments with these respondents and conduct the interviews in the
respective local languages at their homes. Interviews will be tape-recorded. Etc.
56
• Inclusion AND exclusion criteria (who is included and who is not)
• Define sampling strategy (units)
• How many subjects?
(This should fit with the aims and objectives of your study. If you wish to generalize your findings,
you must use a probability sampling technique. Probability samples [simple random, stratified,
cluster, etc.] require a sampling frame. If you are using this, describe your proposed sampling
method, sampling frame, planned and realized sample sizes.
If you have chosen a qualitative approach, make sure you understand the appropriate sampling
techniques in qualitative research [keep in mind that purposive sampling and snowball sampling are
the only sampling techniques permissible for this level of academic research]. If you are planning
secondary analysis of an existing data set, you will have to describe the sampling techniques
employed by the body that collected the data. Relate your sampling to what you wish to do – if, for
example you choose to compare subgroups within your sample [e.g. males and females] make sure
that your samples are drawn accordingly and that you will have enough cases to do the analysis you
wish to do.)
(Here, consider things like consent from participants, approval by the respective institutional review
boards etc
Logistics/budget
Can you do the research within the budget?
WORK PLAN
Describe time span for the study. What will you be doing and when?
-when will the project run? what are the activities at each time
•
LIST OF SOURCES/references
57