UNIT 5 METHODS OF DATA

- -
COLLECTION
Structure
5.0 Objectives
5.1 Introduction
5.2 Levels of MeasurernentrData
5.2.1 Nominal Data
5.2.2 Ordinal Data
5.2.3 Interval Measurement
5.2.4 Dichotomous Data
5.3 Sources of Data
5.3.1 Documentary Sources - Existing Data
5.3.2 Field Sources - New Data
5.3.3 Historical Sources - Existing Data
5.4 Methods of Data Collection
5.4.1 Questioning and Interview Schedule
5.4.2 Observation Technique
5.4.3 Measurement
5.5 Research Tools
5.5.1 Questionnaire
5.5.2 Observation Checklist'Schedule
5.5.3 Observation Sampling Technique
5.6 Procedure for Data Collection
5.7 Let Us Sum Up
5.8 Answers to Check Your Progress

5.0 OBJECTIVES
After reading this unit, you should be able to:
determine the appropriate level of measurement for data in selected studies;
a identify data collection methods;
a 'discoss the data collection instruments;
a explain the procedure for data collection;
a describe the sources of data collection; and
a state the characteristics of research tool in terms of its validity and
reliability.

5.1 INTRODUCTION
Once you have decided on the problem, the research approach and the sample
for your study, you are faced with the task of systematically collecting reliable
and valid evidence upon which you can base your findings. Data (singular
datum) are facts which are observed and measurable phenomena.
The purpose of gathering and summarizing data is to transform them into
information in order to i) identify variableslfacts, ii) measure variables,
iii) describe behaviour, and iv) obtain empirical evidence (i.e. objective, reliable
and valid evidence).
 Methods of Data Collection
5.2 LEVELS OF MEASUREMENTIDATA
Measurement is the process of assigning numbers to the variables. Ways to
assign these numbers include counting, ranking and comparing objects or
events. Measurements, as used in research implies qualification of information,
that is the assigning of some type of number to the data and numbers are then
become the means of computing the information.

The aim of measurement is to provide accurate and objective description of

phenomena that you can communicate to others. Therefore, when the variables
is operationally defined, the investigator must assess the kind of data required
in order to select the appropriate statistical test. From the point of view of
statistical analyses, there are four types of data as discussed below:

5.2.1 Nominal Data

Data are called Nominal Ievel when data can be placed in two or more classes
which are mutually exclusive or exhaustive. When you have nominal level data,
you have categories with different "names" for different groups; e.g. Indian,
American, African. The subject scale value on a nominal variable simply
indicate the group to which he belongs. Putting people into the following
categories would be other examples of Nominal Level Data: 1) boy-girl;
2) student and non-student; 3) Hindu-Muslim-Christian.

Let us see how these would look in tabular form. In the following example
(Table 5.1) we are looking at 1991 Census data on a group of rural population
showing sex of census participants.

5.2.2 Ordinal Data

Ordinal data are discrete or discontinuous consisting of categories of variable
that are ranked, ordered or compared according to a predetermined standard
clothing size, teaching ranks, contest winner, are all ordinal data. A ranking of
patient behaviours according to how often they occur during a given period is
another example. Although there are no equal intervals between points, rank
orders do have a relative order between categories, like first, second, third
ranks or ranked as "always", "mostly", "sometimes", "rarely" and "never".

The interval or ratio data can be converted to ordinal data according to

magnitude of each score, e.g. score, of 81 to 100 may be called 'excellent',
61 to 80 as 'good', 41 to 60 as 'average', 21 to 40 as 'poor' and below 21
as 'very poor'. However, ordinal data cannot be converted to interval or
ratio level data.

Let us see how ordinal data would look in tabular form in Table 5.2. Take
Abde1lah and Levine7s attempt to create a graphic rating for assessing nursing
care. The categories are ranked in order from "care of the highest quality" to
"very poor care" with varying degree of service between.
 IntroductiontoNursing~esearch i) Ratio and ii) Interval Level Data : Ratio and Interval Level data are
the highest level of measurement. They indicate "how much" or the quantity
of a variable. Usually it is a score which represents points on a continuum
and has equal distances or intervals between the points. For example 1,2,3,4
and so on, where there is equal distances between each score (i.e. 2-1 =
3-2 = &3 = 1) and the score can take any point between two numbers
(e.g. 1.1, 1.2, 1.4 etc.); the numbers are continues. These data are superior
because more statistical procedures are available for use with Interval and
Ratio data.

5.2.3 Interval Measurement

It is a scale based on equal units of measurement indicating how much of a
given attribute is present in the subject. It exhibits equal differences ia
amounts. Example : Temperature limitation of this scale is : it fails to measure
absolute magnitude. The scale does not have a true zero. Continues variable
yield interval measurement. Analytical possibilities are greater. The data can be

-
meaningfully added, subtracted and averaged. i.e. If (a = b) and (b = c), then
(a c).

An interval scale is assumed as follows:

a b c d e
1 2 3 4 - 5

The interval from a to c =3-1 = 2. The interval from c to d =4-3 = 1.

We can add this (3-1) + (4-3) = 2+1 = 3. Now note that the interval from
a to d = 4-1 = 3. Expressed in an equation (d-a) = (c-a) + (d-c).

The statistics applicable are:

- can explore mathematical relationship, such as Mean, SD (Standard
Deviation), product moment correlation, Standard Error, Variance etc.
Ratio Measurement
Ratio level is the highest level of measurement. They are distinguished from
internal scales by virtue of having an absolute zero. Ratio scale provides
information regarding : Rank Ordering, interval between object, Absolute
zero.

I Example I
Many physical measures provide ratio scale data - heightlweight, etc.
Since it has an labsolute zero all arithmetic calculations are possible.
Numbers on the scale indicate the actual amounts of the properly being
measured.

The statistic applicable:

All type of statistical procedures are applicable.

The ratio level is distinguished from interval level by having a zero point that
really means zero. In research the distinction between Interval and Ratio
measurement is really not needed because both use the same statistical tests.
Therefore, it is more efficient to group them in one category. Some authors
I
group them as "Measurable" (Continuous data).

The following example shows how ratio or interval level data can be presented
in Table 5.3. For example, if we are looking at the income of villagers by
occupational category might present them in the following table. Here the
income is an interval level data.
5.2.4 Dichotomous Data
- Data that fall into only two categories are called dichotomous data.
- Although these data ordinarily might be considered nominal level (e.g.
pass-fail, qualified-unqualified), they also can be treated as interval level in
many statistical tests because the variable either has an underlying
continues characteristics such as pass-fail or is conceptually changed to
represent the presence or absence of a characteristic.

Levels of data can be measured in various ways, e.g. multi-model

measurement, occurs when several measures instead of just one are selected,
specially when measures are indirect or complicated, such as growth,
compliance. In these cases, indicators are used to measure the phenomena. For
example, height, weight, head circumference and arm measure can be used as
indicators of growth.

The best approach in planning research is to select, if possible, variables

that can be measured at interval and ratio level so that all options are
available to the researcher for statistical analysis. Table 5.4 indicates levels
of measurement.

Note: Although it is haid to get experts agree on cutting points, pulses below 50lm and above 150lmin are
not considered normal and so are marked 'X' with nominal level column.
lntro*uction to Nursing Research You can see that here the categories have no relative standing or order among
them like ordinal data: This is the lowest level measurement. Ratio, interval
and ordinal data can be converted to nominal level, however nominal level data
cannot be raised to higher levels.

The sources a student should tap for collection of data for a research
study vary with hisher interests and the type of study. However, these
can be classified as : i) docurnentary sources, ii) field sources, and
iii) historical data.

5.3.1 Documentary Sources

Those information which are contained in the published and unpublished
documents, reports, statistics, manuscripts, letters, diaries and so on, are
grouped under two broad categories, i.e, primary sources and secondary,
sources.
Primary sources provide data which are gathered at first hand where the
responsibility and promulgation remain under the same authority that
originally gathered them. Data from primary sources may be gathered by
observation, personal interview, conference, correspondence, questionnaire
and other devices,
Secondary sources provide data that have been transcribed or compiled
from original sources; here the promulgating authority is different from the
one which controlled the collection of data at first hand, e.g. the Census of
India is generally cited as the primary source whereas the Health Statistics
on India based on Census of India are examples of the Secondary Source.
Note that promulgation means publication, dissemination, making known to
public or exposing to public.

Documentary
(Documents, I
reports, statistkls,
letters, diaries,
etc.)
M d $owces
pw-,
I

-
opinion, attitude,
Compet~nce,
Envirmmt
Evmts
I

liistorical D&I
Docwnents,
tnakwhts ofC%S~MJ

history, -1
sources, artkcts,
etc.

II
Primary (gathered Secondary
first hand) e.g. (transcribed or
Decennial compiled from
population census original source)
Sample e.g. Health
registration system statistics
Survey of cause of compiled by
death CHEB
e Family Welfare
statistics

72 Fig. 5.1: Source of Data

 Methods of Data Collection
5.3.2 Field Sources
Include living persons who have the knowledge about or have been in intimate
contact with social conditions and changes over a considerable time. These
People are in a position to describe not only the existing state of affairs (e.g.
patient tells about his conditions) but also the observable trends (e.g. see how
the 1CU nurse handles modem gadgets) and significant milestones in the
society. Field sources also include conditions, environment and events that
are observable and measurable, such as sanitary conditions, humidity.
Most survey studies and all experimental studies use field sources for data
collection.

5.3.3 Historical Sources

Historical Data are a combination of documentary and field sources. These
consist of documents and various historical sources to which the historian
himself has access. These may include: materials of cultural history and
analytical history, personal sources of authentic observers and witnesses, and
these may consist of oral, written evidences, artifacts, etc. An example ii,
M.Phil thesis under. Delhi University, 1989 on the development of collegiate
programme in nursing in Assam contains Government documents, university
minutes, TNAI report, interview with first Principal, Health Secretary, etc.
which indicate "who proposed", "who supported and "who assisted" in the
development of the institution.

5.4 METHODS OF DATA COLLECTION

The various method; of data collection are discussed in the following text.
 Introduction to Nursing Research 5.4.1 Questioning and Interview Schedule
Meaning: The questioning is a technique of data collection where the
sources of data is the subject himself/herself. Here, the data are collected
through self reporting which requires some form of quest~oningto obtain
needed information. Questioning of subjects in a research project generally
is accomplished through the use of a formal instrument, which are:
Interview Schedule, Questionnaire or Opinionnaire.
Interview schedule is an oral questionnaire. Here the researcher
(interviewer) orally asks questions to the participant (interviewee) of the
study, either in face to face meeting or over the telephone or video
recording. Interviewing is one of the questioning techniques in which the
researcher has an ability to control the level of questioning. Interview
Schedule is a questionnaire (more structured) that is read to the respondents
by the interviewer whereas an interview guide (less structured) provides
ideas but allows the interviewer a freedom to pursue topics in depth.

Questionnaire comprises a series of questions prepared by the researcher that

are answered and filled in by all the respondents. This is usually self-
administered. The opinionnaire is a type of questionnaire which comprises of a
series of questions that may refer to matters of opinions to be answered by the
respondents.

The interview schedule or the questionnaire the researcher uses different forms
of questions, such as closed-ended (or fixed alternative), open-ended or
partially close-ended questions.

Example: Close-ended items

Exampfe f
Have you ever taken a course in statistics?
*I)
( 1
( 1 No
b 2) How important is it to you to work in an institution that has ample
opportunities for a d v a n c e ~ n t ?
( ) extremely important
( 1 very important
( ) somewhat important
( ) not at all important

Here the questions offer respondents a number of alternatives, such as 'yes' or

'no', a short response or check on item from list of suggested responses as the
item 2 in the Example 1 above shows. These are close-ended items.

Example: Open-ended items

Example 2
1) Why did yuu choose to take your graduate work at this university?
2) In your opbion what are the causes for cancer? 1

In the above example the questions have no answer choices from which
respondents select their response. Instead the respondents must "&eateWtheir
74 answers and respond to the questions in their own words.
Example: Partially Close-ended Questions Methods of Data Collection

For what reasons did you retire before the age of

ii) Desire for more free time

iii) Was assured of more than adequate in
1 iv) List as many other reasons as you can thin 1
In Example 3 although answer choices are provided, the respondent has the
option of creating hislher own responses as well (see Choice IV in
Example 3).

5.4.2 Observation Technique

What does observation technique mean: Observation Is one of the basic
and oldest research methods to gather data. Observation stops being a
normal part of daily life and becomes a research method if it is
systematically planned and recorded and when both observation and
recording are checked for their validity and reliability. It is a technique for
collecting the data through occurrences that can be observed through senses
with or without mechanical devices.

Observation is a two part process i.e. i) someone is observing--observer,

and ii) there is something to observe-the observed. It is primarily centered
arowd naturalistic conditions, termed as field research, i.e. the observation
occurs in the natural environment of subjects. Laboratory setting is used
when target behaviours are observed in environments different from the
environment in which the behaviour normally occus. It usually occurs in a
clinic or laboratory setting specially equipped for unobtrusive visual and
auditory monitoring.

In observation there are four broad questions that confront the researcher. These
are:
a) What should be observed?

b) How should observation be recorded?

c) What procedures should be used to try to assure the accuracy of

observation?
d) What relationship should exist between the observer and the observed,
and how can such relationship be established?

What is to be observed: Phenomena that can be observed. 'There are various

phenomena that can be observed in nursing research. For example:
Characteristics and conditions of individuals, such as physiological
conditions.
Verbal communication behaviour, such as linguistic behaviour, people's
conversation.
a Non-verbal communication behaviour, such as, facial expressions, touch,
body movements.
Activities, for example actions that serve as an index of health status,
performance activity of nurses: performance of procedure.
Environmental characteristics, such as noise levels, cleanliness that have
profound effect on health or individual's behaviour.
Introduction to Nursing Research Types of Observation
Types of observation relate to the observer-observed relationship. It has been
described as "concealment and participant", "concealment and non-
participant", "no-concealment and participant" and "no-concealment and
non-participant" as diagrammatically shown in Figure 5.4.

No Canmalment and Participant

Observer

~ ~ n c e a l r n e aria
Non-parttapant
Observer behlnd
nt
+
[U3
Observer-
+--- Non-concealment and
Non-participant
screen (No screen)

T
Fig. 5.4: Concealment and participant

"Non-concealment" and "Concealment" refer to whether the observer makes his

intentions known to the persons observed or not, whereas "participant" and
"non-participant" refer to the role of observer; that is, whether the observer is a
part of the social setting or makes the observation from periphery of a social
setting.

The instruments used for the technique can be unstructured e.g. logs and field
notes, Anecdotes or can be structured e.g. "Logs and Field Notes": Logs is a
record of events and conversations and usually is maintained on a daily basis
by field worker.

"Field notes" may include daily log but tend to be much broadel; more
analytic, and include more intelpretation than more listing of
occurrences.
"Anecdotes": Anecdotes focus on behaviour of particular interest. Anecdote
typically selects specific kind of events and behaviours for observation
beforehand. The observer objectively, and accurately records the
observation (Refer Block 2).
"Checklists": It is a tool used for structured observation where
phenomena are recorded by putting a tally against a particular behaviour,
event or characteristic depending upon its presence or absence (Refer
Block 2).

When developing a checklist researcher must list all the expected behaviours
related to the vaiiable being measured. So that alr relevant behaviours in the
subjects will correctly' be identified.

5.4.3 Measurement
The third method of data collection is measurement where standardized
instruments are used to measure variables. For example, scientific
instruments such as thermometer, barometer, weighing scale are instruments for
physical measures of temperature, pressure and weight respectively. Precision
measure depends on refinement of the instrument. A standard instrument
should be used and preferably: the same instrument should be used: it is for
measuring all subjects.
 Methods of Data Collection

What is the difference between "measurement" and "assessment"? These two

terms are sometimes used synonymously. But there is a difference between the
two. In a simple way it can be said that measurement means to ascertain the
dimensions, quantity or capacity of objects or events, whereas assessment can
be defined as a systematic procedure of collecting quantitative as well as
qualitative data to describe behaviour or characteristics. For example 8 rating
scale to survey job satisfaction of nurses or survey learning needs of nurses
qualifies as assessment where as a knowledge test on AIDS is used to measure
knowledge of AIDS in students.

In measurement, numeral are assigned to objects or events according to rule to

give a quantitative meaning. A numeral is a symbol of the form 1, 2, 3 and so
on or I, 11, I11 and so on. When quantitative meaning is given to the numerals,
these become numbers. Hence, number is a numeral that has been given a
quantitative meaning. It is easy to understand measurements used in natural
sciences, which measure, length, weight and volume. Accurate scales are
available for these. Measurements used in psychological, sociological,
educational characteristics or nursing care standards are difficult to understand
unless assignment of numerals to objects or events are made according to
some accepted rules set up on rational or theoretical basis. For these
measurements scales are prepared, like aptitude scale, nursing standard scale,
etc.

A scale is a technique that assigns a numerical score in order to place

individuals along a continuum with respect to the attribute (or characteristic,
phenomenon) being measured. It helps us to distinguish among individuals
(or things) in terms of the degree to which they possess a specific
attribute.

Types of Measurement
The types can be classified under three categories:

Physiological measurement

. Psychological measurement

Sociological measurement

Physiological ~easurementcan be classified as In Vivo and In Vitro.

In -Vivo includes measurements of those which are performed directly within or
on living organisms themselves; such as blood flow determination through
radiography, IVP, Cholecystography.

An In Vitro measurement by contrast is performed outside the organism's

body, such as blood glucose, blood urea, serum Na', K+.
,
Physiological measurements can also be classified as listed below:
Physical measurement - Temperature, Volume, Pressure, Electrical
measures.
Introduction to Nursing Research 8 Chemical measurements - hormone levels, sugar, urea, Hb%, potassium
level, etc.
8 Microbiological measurement -- bacterial connts and identification.
8 Anatomical and Cytological measurements - X-rays, tissue biopsies,
tomography, CAT Scan, etc.

Psychological Measurement mainly makes use of psychological tests and

inventories. Psychological test is an instrument designed to measure only a
certain aspect of human behaviour. These tests yield objective and
standardized descriptions of behaviour, quantified by numerical scores.
These tools attempt to quantify the feelings, attitudes, opinion of individuals,
etc, for example, measurement of attitude, personality, intelligence, and
achievements.

Sociological Measurement attempts to measure groups or individuals in

relation to a social set up. For example, Social Distance Scale which attempts
to measure to what degree an individual or group or Individuals is accepted or
rejected by another individual or group. The first "feeling" reactions or needs
are checked without rationalizing. Socio-economic Status Scale such as SES
Scale (Urban) B. Kuppuswamy, are the few examples.

5.5 RESEARCH TOOLS

Questionnaire and interviews are probably the most frequently used data
collection methods in rlursing research. Observation is also an important
method of testing research hypothesis and seeking answers to research
questions. In addition, various physiology and psychological measures are used
by nurse researcher.

One of the sources of data is subject himseif or herself when the data
collection procedure requires self-reporting through some form of
questioning either a questionilaire or an interview schedule can be used for
the same.

5.5.2 Questionnaire
Questionnaire is a paper and pencil. self reporting instrument can be used to
measure knowledge levels, opinions, attitudes, beliefs, ideas, feelings and
perceptions, as well as gather factual information about the respondents.

Questionnaire is the most common research instrument and is comprised a

series of questions that are answered by all the participants in a sanlple.

Interview Schedule: Intemieui schedule, in a sense is an oral questionnaire.

Here researcher (interviewer) orally asks questions to the participant (or
interviewee) in the study, either in face to face meeting or over the telephone/
video recording.

Interviewing is the one questioning technique in which the researcher has an

ability to control the level of questioning.

Interview Schedule is a questionnaire that is read to the respondents by the

researcher (interviewer whereas an interview guide provides ideas but allows
the interviewer fieedom to pursue topics in depth.)
Advantages and disadvantages of both questionnaire and interview are listed Methods o f Data Collection
s
below-:

2. Respondents may omit any item

data from a widely scattered without giving any explanation.

information that

availability of time and interest of

no or low literacy cannot

responses. - answer.
8. Analysis and interpretation of 8. If respondent is promised
data can be easily accomplished anonymity, it is impossible io
when structured. know who returned the question-
naire in case follow-up is needed.
I
Interview %

1 Data from each interview are I. It.is time consuming and costly,
usable, whereas this may not be when it involres travelling.
true for each questionnaire returned.
2. Indepth data can be obtained, 2. Rapport ahd interpersonal
since researcher can pursue any relationship are important aspects
question of special interest. technique. Therefore, element
often can b e from both
interviewe! and interviewee.
3. Respondents are less likely to 3. It may be difficult to make a
give "don't know" responses comparison of data collected by
or leave a question unanswered. interviewet with another unless a
Introduction tv Nursing Research

Qpes and Structure

The design of a useful schedule normally involved extensive planning,
consultation with colleagues and experts, pretesting and revising. Some of the
major considerations that to be kept in mind when constructing a questionnaire
andlor interview schedule are:
i) Form of Schedule
Schedules vary in extent to which they are structured. These can be structured,
seini-structured, unstructured. Majority of the interview schedule and
questionnaire fall in between these two extremes of rigid standardization
(structured) and total absence of structure (unstructured). That is, schedules are
most often composed of a combination of questions which vary in their degree
of structure. In general, questionnaire tend to be more standardized than
interview schedule.
ii) Form of Question
Review the forms of question which could be close-ended, open-ended and
partially close-ended questions.
Some advantages and disadvantages of open and close-ended question are listed Methods of Data Collection
in Fig. 5.4. The partially close-ended questions provide a compromise. Although
answer choices are provided, respondents have option of creating their own
responses also.

Close-Ended
I
a Easy to respond Limited to specific topic to be
answered on a specific way.
Therefore sometimes data is
superficial or lacks depth.

1I a Responses are easy to

summarize and tabulate.
I I
I--- -
/ Analysis of data are suited to I I
I many statistical analysis. I I
Open-ended %

I
Stimulate free thought, solicit a Responses usually involve
suggestions and probe people's recalling of past experiences.
memories. Therefore there may be element
of error or bias.

a Indispensable for exploratory All respondents may not give

studies in which researcher's articulating answers.
main purpose is to find out the I

most salient aspects of a topic.

1 a Not suited for mailed question. 1

a Difficult to have statistical
analysis and analysis is time
consuming?

Fig. 5.4: Open- and Close-ended Questions- Advantages and Disadvantages

iii) Question Content

The content of questions will naturally be guided by the aims and needs of the
researcher. Direct questioning is particularly useful for obtaining information on
knowledge, beliefs and practices. In deciding on the question content. the
researcher should not lose sight of initial goals and objectives of the project. as
it is not uncommon to hate a schedule get out-of-hand in terms of its length,
coverage and complexity.
iv) Question Wording
Number of considerations that should be h e . in mind are :
a Clarity
Ability of respondents to reply
a Bias
Handling sensitive or personal information (will be dealt in detail in the
section on writing good questions).
v) Consideration for Responses
Few general considerations to be kept in mind are:
a) Coverage of alternatives - responses should adequately encompass all
of significant alternatives.
Introduction to Nursing Research b) Overlapping responses - should be avoided. The alternatives should
be mutually exclusive e.g. How many children do you have?
Faulty item Correct

( ) None ( ) None

( ) 5 or more ( ) 5 or more

c) Ordering responses - There should be some underlying rational for

the ordering which alternatives are presented to the respondents.

d) Response Length - The response alternatives should not be too lengthy

since it is inefficient and cumbersome for the subjects to read detailed
replies.
vi) Organization
In organization of the schedule the researcher must plan on the questioning
sequence (order of questions) and the schedule format.

Question Sequence (Order): Some thoughts should be given to the sequencing

of the questions so as to arrive at an order that is psychologically meaningful
to the response and encourages them to answer all questions. The order in
which questions are asked affect the responses as 'well as the overall data
collection activity. Responses given by subjects will tend to be more accurate
and representative of the sample when certain principles are
followed:
Every item in a questionnaire should relate to the topic under study.
Items should be organized into units.
General questions should lead to specific ones.
One questions should not influence the other.
Personal background information such as age, educational level, marital status
etc. usually included in the questionnaire either in the beginning or end.

Schedule Format (Lay-Out): The way a questionnairelinterview schedule is

laid may affect response rate and the accuracy of the responses given. Formats
that are easy to fit and pleasing to the eye are usually completed by the
respondents.

5.5.2 Observation Checklist/Schedule

This section Refer to the section on observation technique given in subsection which
describes the phenomenon that can be observed and types of observations.

i) Definition
Observation is a technique for colleCting data through occurrences that can be
observed through senses with or without mechanical devices. There are two
parts, i.e. someone is observing or observer and there is something to be
observed.

There are different instruments that can be used for collecting data through
observation, this section we will be primarily describing development of
observation schedule and checklists. .. .
ii) Advantages and Disadvantages Methods of Data Collection
Introduction to Nursing Research 5.5.3 Observation Sampling Technique
Since it is usually impossible to observe behaviour for extended period of time
because of fatigue and boredom one must plan how and when to make the
observalion. The most commonly used techniques are:
i) Time Sampling Method: This procedure involves the selection of time
periods during which the observation will take place. One decides the day
into units that are appropriate for observation. The time frame may be
systematically selected e.g. one can observe intermittently at 10 minutes
interval for 8 hours a day or intermittent instantaneous observation may be
carried out at 10 minutes interval for 24 hours. Time frames can be
selected at random. Here it is important that the researcher uses some
system of signal when the observer should start looking and when to stop.
Duration of observation should be decided before hand.
ii) Event Sampling: When we want to observe an entire event in order to
give the subject the opportunity to perform all of the expected behaviours.
The observer selects integral behaviour or event of a pre-specified type
of observation. Researcher either requires to know the occurrence of event
or be in a position to wait for their occurrence e.g. a researcher observed
aseptic practices in urinary catheterization where s h e observed the full
procedure against the steps listed under different areas of procedure e.g.
washing and drying hands, wearing gloves, etc. Slhe carried out
observation in surgical ward, surgical O.P.D., casualty department.

5.6 PROCEDURE FOR DATA COLLECTION

The planning for research continues till a pilot study is conducted to see that
the whole research design works. The next step is the stage of implementation.
Before we actually act into the process of data collection, it is necessary to
make a meticulous plan, so that data collection process is completed smoothly.

By now the instrument is prepared and it is necessary to find out i) what data
are to be collected? ii) From whom are the data to be collected? iii) Who will
collect the data? iv) From where will the data be collected? And v) When is
the data to be collected?

i) What data are to be collected

At this stage the researcher need to consult statistician to make sure that
the type or the form of data that s h e gets would match data analysis plan.
For example, certain statistical test can only be done if the level of data is
interval or ratio. In this case the data has to be intervallratio level.
ii) From whom are the data to be collected?
This involves consideration on sample subjects. For example, who are the
study subjects? How are they selected? How many are they? That is, the
researcher needs to list the criteria for selection of the sample.
In cities and cosmopolitan areas, institutions have established committees
which review research proposals to assure the rights of human subjects. In
fact one may call it th'e Committee on Human Rights. And, some
governm~ntswill not sponsor research activities unless and until the rights
of subjects are shown to be protected. This degree of development has not
reached many villages; however, all plans for conducting research should
be cleared with government and/or village authorities. Usually, a detailed
project proposal is required indicating the nature of the study, the subjects
involved, the kind of independent variable or treatment which will be
introduced and by whom data would be collected. All activities,
procedures, treatments, and/or methods must be delineated so that
 authorities can assess the study from a human rights point of view. This Methods of Data Collection
not only protects the villager, but protects the investigator as well.
iii) Who will collect the data?
One of the common questions asked is "who will collect the data'!" The
researcher andor hislher team is solely responsible for collection of data. If
you are going to collect data yourself, it is important to try out the
procedure on a small sample on similar subjects. This is called pre-testing,
and is often carried out during a pilot phase of research process. Pre-testing
is a step in the development of the instrument. Besides its function in
establishing reliability of the tool, it also helps the researcher in finding out
whether the data collection procedure is feasible.

If you have people collecting data for you they need to be advised on how to
answer questions honestly and use only the methods and equipment agreed
upon by the villagers and the leaders. The integrity of the people collecting the
data is a fundamental element of success.

A rigorous training programme has to be set up. First, you explain the
collection procedure, approach to subject and method of recording the data.
Then you accompany them when they collected data and simultaneously collect
the same data; tally both recordings. Training is not complete till you find them
capable of collecting data accurately and completely. In the case of observation
techniques, the final data are always more reliable when activities are observed
and recorded independently by two observers simultaneously.

Submission for data collection is to be obtained for all studies from the head of
the institution or village panchayat or as the case may be. An example of a
letter is given below.

Contents Remarks
From. .......... Enter your official address
To
The Medical officer Accurately address theahead of
'X' Health Centre institution/organization, including titles
'Y' District and initials
'Z' Country
'Address by name if it is known to you.
Date. ..............

Dear Dr. X
-

I am conducting a study under the Give your project title and the
auspices of Kanke Primary Health purpose of the study. Be concise,
Centre to investigate the crisp, to-the-point.
"(Title of Research)". It is hoped
that the information obtained from
this study will be useful for the
State. I therefore request you and
your staff to participate in the study.
-

In this study, a health staff member Roughly define your population.

is one who is employed full-time for
providing health care. Since the valid
data necessitate the participation of
entire health staff as defined above.
I would ask you to request your
staff to agree to such participation.
- 85

4-
 - -

Introduction to Nursing Research

Participants will be required to Assure anonymity and confident
complete a questionnaire. Anonymity state the nature of the participant
of participating health centres and and approximate time commitment
individuals will be protected. Your referred by the participants.
participation in this study, I may add
here, would involve the completion
of: a) A questionnaire; and
b) an interview with me, both of
which are not expected to take more
than 1 to 1-112 hours of your time.
--

May I pleas2 have your permission Listing the population is necessary

to conduct this study in the ........ and a must if you want to do
Health Centre? I also request you random sampling.
to send nie the names of the staff,
so that I can mail the questionnaire
directly to the participants.
I will appreciate your consideratiog It is a courtesy to send the
of this request at the earliest expenses involved - it shows you
possible date. For your convenience, mean business.
a reply'sheet and return postage
stamps are enclosed.
Thanking you for your cooperation. Be gracious.

Yours sincerely,

iv) From where will you collect the data?

This refers to the setting depending on your problem for investigation, you
need to collect data about the location, road communication, climate,
language spoken, food habits, religion, etc. If your research is an
experimental study, make sure the control group subjects and experimental
group subjects do not mix.
v) When are the data to be collected?
In developing countries, the 'when' of data collection is a major
consideration. In one study, data collection took place during the harvest
season. Each participant had to be sought out at evening time because all
men in the fields during the study. The follow-up could not be conducted
because two months later, the entire village was flooded during the
monsoon.
Activity 1 -
Read the research article "Qualified Personnel or Traditional Midwives
women's choice in Rural India" (TND-HND-007) and
a) Identify the instrument used.
b) State the forms of question used:
c) Describe the partslsection of the instrument and the forms of data
collected.
d) State why the pilot study was done?
Methods of Data Collection
Introduction to Nursing Research Type of Data Justification
i) Column (11) ....................................................
ii) Column (111)
iii) Column (IV)

5.7 LET US SUM UP

In this unit you have learnt about different levels of measurements that is
nominal, ordinal, interval, ratio and dichotomous data. The two types of data
source i.e. primary and secondary. This unit also describes the most essential
part of research study that is method of data collection that is questioning,
observation schedule and measurement. The researcher has to decide on the
methods of data collection and tool b a e d on its problems and purpose of
study. The research tools commonly used are questionnaire, interview schedule,
observation check list, etc. methods of direct questioning are indispensable as a
meails of collecting data on human subjects, but are susceptible to errors of
reporting. Observation methods are technique for acquiring data through direct
observation of phenomenon.
TerminaVSelf Assessment Exercises
1) Give three advantages of observation as a data collection technique.
2) List ten criteria of good question.
3) What are the different ways of recording interview responses.

5.8 ANSWERS TO CHECK YOUR PROGRESS

Check Your Progress

1) i) d ii) b iii) a iv) d

2) a) Questioning
b) Observation.
c) Measurement
3) i) lnterval Continuous score indicating blood sugar in mg1100 rnl or Ratio
Level of blood. Score can take any point between two numerals.
ii) Ordinal Ranked data showing first to tenth rank. There are no equal
intervals but have a relative order between the categories.
iii) Nominal Data are placed in two classes i.e. normal and above normal,
which are mutually exclusive.
Activity 1
a) Instrument Used: An interview Schedule
b) Forms of questions used : Structured and Open-ended questions
c) i) PartsISection of the instrument
First Section : The background data including religion, age, income,
occupation, education, size and type of family.
Second Section: Items on the factors related to the kind? and nature
of services available to the mother, her personal preference for trained
or untrained personnel, the desirable characteristics of the personnel and
the influence of the family in !he selecting the type of personnel.
ii) The forms of data collected. Opinion data, e.g preference or choice of
utilizatih.
4
d) Pilot Study: To test the feasibility of the study and to make necessary
changes in the instrument.
TerminaVSelf Assessment Exercises Methods of Data Collegtio

1) i) Relatively simple, quick and inexpensive method.

ii) Researcher can gather data from a widely scattered sample.
iii) Respondents can remain anonymous.
iv) Easy tool to test reliability and validity when structured.
v) Can find her own time to respond to each question.
Ten Criteria of good questions:
i) Use simple and direct words.
ii) Questions should be clearly stated.
iii) Complicated or confused terms should be clearly stated. ,

iv) Do not use double barrelled questions.

v) Do not use double negatives.
vi) For putting special emphasis, underline words.
vii)Avoid leadin$ questions.
viii) Check that alternative responses are appropriate.
ix) Avoid using 'you' in questions.
x) Either omit or explain why personal questions are asked.
3) Different ways of recording interview:
i) Simultaneous recording in long hand or short hand.
ii) Use of audio or video tapes.