Chapter 2 • historical documents,

• statistical reports,
Methods of Data Collection and Presentation
• records of institutions and
• other sources
2.1 Methods of data collection
Source of Data: Self-administered questionnaire : is a method
Statistical data may be obtained from two of data collection in which researchers can give
sources, namely, primary questionnaires with instructions directly to
respondents or mail them to respondents who
and secondary read instructions and questions, then record
1 Primary data: their answers and give it back or return it by mail
again to data collecting agency
data measured or collected by the investigator
or the user directly from the source. Primary Advantage:
sources are sources that can supply first hand • It is the cheapest and can be conducted
information for immediate user. by a single researcher.
2 Secondary data: • A researcher can send questioners to a
wide geographical area.
When an investigator uses data, which have • The respondent can complete the
already been collected by others, such data are questionnaire when it is convenient and
called secondary data. Data gathered or check personnel records if necessary.
compiled from published and unpublished
• Mail questionnaire offer anonymity and
sources
avoid interviewer bias.
• They are very effective, and response
We distinguish three basically different methods rates may be high for a target population
of collecting data. These are; that is well educated or has a strong
interest in the topic
1 Extraction of data from records
Disadvantage
2 Self-administered questionnaire
• A low response rate is the biggest
3. Direct investigation-measurement
problem.
(observation) of the subject and interviewing
• A researcher cannot control the
(face-to-face, telephone) our first step is to
conditions under which a mail
decide on which of these three methods to use
questionnaire is completed.
• Researcher cannot visually observe the
Extractions of data from Records respondent’s reactions to questions,
physical characteristics, or settings.
A mass of information about the population
studied by social surveys is available in
 • Mail questionnaire is not suitable for Interviewing (face-to-face, telephone)
illiterate community.
• Face-to-face interview is a social process
that involves the interviewer and
Direct investigation - measurement respondent.
• It is the process in which the interviewer
These are:
meets the respondents, explains the
• Observation and purpose of the study, forwards a set of
• Interviewing (face-to-face, questions and records the answers.
telephone) • It is widely used in economic and social
surveys
Measurement or Observations
Information on a topic can be gathered by
Advantages of face-to-face interviews
measurement if it is physically measurable or
observable. Common types of data collected o Have the highest response rate and
by observation and measurement include: permit the longest questionnaires.
o Interviewers control the sequence of
• Land area measurement
questions and use some probes.
• Crop output measurement
o Respondent is likely to answer all the
• Anthropomorphic measurements questions alone.
• Animal weight gain o Interviewers also can observe the
• Instrument recording or readings surroundings and can use nonverbal
(e.g. rainfall, temperature, etc) communication and visual aids.
• Physical measurement or o Well-trained interviewers can ask all
examination of people types of questions including complex
• Counts of human, animal and plant questions.
populations
• Direct observations of work
Methods of data Presentation
• Exchange activities (e.g. purchases
and sale prices). Textual Method: – a narrative description of the
data gathered
Data collection by measurement can be
undertaken is several ways, some of these are: Tabular Method or frequency distribution :– a
systematic arrangement of information into
• The direct measurement of a physical
columns and rows
characteristic using an instrument
• The observation of people engaged in an Graphical Method :– an illustrative description
activity; and of the data
• Recording of relevant aspects of their
activities
 Categorical Frequency Distribution:
THE FREQUENCY DISTRIBUTION TABLE (FDT) • Data are classified according to non-
numerical categories.
A FDT is a statistical table showing the
• Categories must be mutually exclusive and
frequency or number of observations contained
exhaustive.
in each of the defined classes or categories.
• Used to present nominal and ordinal data
Parts of a frequency distribution table:

• Heading
Nominal data: Here the construction is straight
• Body
forward: count the occurrences in each category
• Stubs or classes
and find the totals.
• Caption
Example:The martial status of 60 adults
Frequency distribution: is a basic techniques
classified as single, married, divorced and
that provide rich insights into the data and lay
widowed is presented in a FD as below:
the foundation for more advanced analysis.
Marital Single Married Divorced Widowed Total
A frequency distribution table: lists categories Status
of scores along with their corresponding Frequency 25 20 8 7 60

frequencies.
Ordinal data: The construction is identical to the
Frequency distribution nominal case. However, the categories should
be put in an ordered manner
It is a grouping of all the (numerical)
observations into intervals or classes together Example: Satisfaction of hospital admission in
with a count of the number of observations that Gondar hospital size of 80 is presented a FD as
fall in each interval or class. shown below.
Satisfaction V.Satisfied Satisfied Dissatisfied V.Dissatisfied Total
A frequency distribution has two main parts: Frequency 15 36 3 7 80

• The values of the variable (if

quantitative) or the categories (if Numerical Frequency Distribution
qualitative), and
o data are classified according to numerical
• The number of observations (frequency)
size.
corresponding to the values or
o used to summarize interval and ratio data.
categories.
o may be discrete or continuous, depending
on whether the variable is discrete or
continuous Population
There are two types of Frequency distributions
1. Categorical (or qualitative)
2. Numerical (or quantitative)
1. Discrete (Ungrouped) Frequency Distribution interval and adding 0.5 to the upper limit
interval.
Count the number of times each possible value
is repeated
Example:In a survey of 30 families, the number Continuous or Group Frequency Distribution
of children per family was recorded and
Class Mark (x) – is the middle value or the
obtained the following data:
midpoint of a class interval. It is obtained by
424328344228534545435273367 getting the average of the lower class limit and
384 the upper class limit
These individual observations can be arranged in Class Size or width (W) – is the difference
ascending order of magnitude to from an array: between the upper class boundary and the
lower class boundary of a class interval
222223333333444444445555677
888 Relative Frequency (RF) – these are the
percentage distribution in every class interval.
The distribution of children in 30 families would
be: Class Frequency – it refers to the number of
observations belonging to a class interval, or the
Number 2 3 4 5 6 7 8 Total
number of items within a category
of
children
Number 5 7 8 4 1 2 3 30 Steps in Constructing a (FDT)
of
family Example: The blood glucose level for 50 patients
is shown below.

Continuous frequency distribution Construct a frequency distribution for the

following data
The purpose of this kind of a table is to select a
set of intervals on the number line, then count 51 65 68 87 76 56 69 75 89 80 61 66 73 86 79 70
the number of values that fall into each interval 71 54 87 78 68 74 66 88 77 67 73 64 90 77 72 52
67 86 79 74 59 70 89 85 55 63 74 82 84 57 68 72
Definitions of some basic terms 81 83
Range (R) – the difference between the highest Step 1: Find highest and lowest value and
score and the lowest score. compute range R, using the formula:
Class Interval (k)– a grouping or category R = HighestScore–LowestScore (1)
defined by a lower limit and an upper limit.
Step 2: Compute for the number of class
Class Boundaries (CB) – these are also known as intervals, K, by using the formula Decide k with
the exact limits, and can be obtained by the help of Sturge’s rule
subtracting 0.5 from the lower limit of an
k = 1 + 3.3 logn (2)
Note: The ideal number of class intervals should Step 2: the class limit using Sturge’s rule
be 5 to 15. Less than 8 intervals are
k = 1 + 3.3 log 50, k = 8 (5)
recommended for a data with less than 50
observations/values. For a data with 50 to 100 Step 3: Find the class width
observations/values, the suggested number
w = r/k,w = 39/8,w = 4.875 ≈ 5 (6)
should be greater than 8
Step 4: Select the starting observation as lowest
class limit and Add the width to that observation
Example: The blood glucose level for 50 patients
to get the lower limit of the next class.
is shown below.
Keep adding until there are 5 classes 51, 56, 61,
Construct a frequency distribution for the
66, 71, 76, 81, 86, 91
following data
Step 5: Find the upper class limit; e.g.
Step 3: Find the class width; w=R/k and closest
ones (rounding up) thefirstupperclass = 51 − (U = 56 − 1) = 55 (7)
w = R/k (3) 55, 60, 65, 70, 75, 80, 85 and 90 will be the upper
class limit
Step 4: Select the starting observation as lowest
class limit (this is usually the lowest
observation). Add the width to that observation
So combining step 4 and step 5, one can
to get the lower limit of the next class. Keep
construct the following classes: So combining
adding until there are k classes
step 5 and step 6, one can construct the
Step 5: Find the upper class limit following classes.
Step 7: Find the class boundaries by subtracting class limit
0.5 from each lower class limit and adding 0.5 to
51-55
the UCL as shown.
56-60
Step 8:Tally the data
61-65
Step 9: Write the numeric values for the
frequency column 66-70
Step 10: Find cumulative frequency. 71-75
Step 11: Find relative frequency and /or relative 76-80
cumulative
81-85
86-90
Solution of Example:
Step 6: Find the class boundaries by subtracting
Step 1: Highest value 90 and lowest 51 then
0.5 from each lower class limit and adding 0.5 to
range = 90 − 51,range = 39 (4) the UCL as shown Example: