Chapter 3 & Likert Scale Reviewer
Chapter 3 & Likert Scale Reviewer
Chapter 3 & Likert Scale Reviewer
Introduction
O Chapter 3 or Methods of Research is all about how you will conduct
your research. This includes your research design, number of samples,
your research questionnaire or data collecting instrument (DCI), data
gathering procedure, analysis and others.
Research Design
O The Research Design should provide for the scheme that will be
used in collecting, analyzing, and interpreting data.
O A descriptive type of research design is always been used in a thesis or
dissertation.
O Descriptive research answers the questions how, why, how,
what, when and where.
The Research Design
O At the minimum, it should include
O (1) a specification of the variables to be employed,
O (2) a detailed statement of how observations (i.e., data collection) are
to be organized and conducted,
O (3) a discussion of how the data collected will be analyzed, and
O (4) a justification of the methodology that was chosen.
Sample Size and Sampling Design
O This part includes the proposed total number of samples to be used as
respondents of the research.
O It also includes the total number of population of the proposed
research.
Data Gathering Procedure
O This part includes the way in which the researcher will collect his/her
data i.e. through survey, focus group discussion or through an in-depth
interview.
Data Analysis
O This part includes the use of tables, charts and graphs to be used and
mathematical tools such as weighted mean, standard deviation, chi-
square test, fishers test, Kruskall Wallis test, Spearman test, t-test,
ANOVA and Pearson-r test.
Data Analysis
O Standard deviation from the mean is used as a measure of variability
when the measure of central tendency is the mean.
O Chi-squared test to determine the relationship between variables.
O Fishers test - to compare two unpaired group
O Kruskall Wallis test to compare 3 or more groups
O Spearman test to compare association between variables
O T-test to compare one or two groups
O ANOVA to compare 3 or more groups
O Pearson-r test to compare the association between variables.
O The most frequently used measures of variability are the range,
standard deviation and the variance.
O Range is defined as the distance between the highest and the lowest
score. (H-L=R)
O The Standard Deviation from the mean is used as a measure of
variability when the measure of central tendency is the mean.
O It is this pair of statistics that is always used for interpretation of data.
O Variance is the square of the standard deviation.
Appendices includes:
O References
O Books, Journals, Periodicals, Internet:
O Appendix I Data Collection Instrument or Questionnaire
O Curriculum Vitae
Data Collecting Instrument (Research Question)
O Refers to the device used to collect data, such as a paper
questionnaire or computer assisted interviewing system
Data Collection Instrument (DCI) or Questionnaire
O This part includes the proposed instrument or questionnaire to be used
for the respondents
O The proposed instrument or questionnaire should be included at the
last part of the paper.
O The data collecting instrument should have a minimum of 20 questions
for the quantitative or 10-15 questions for the qualitative type.
O Monitoring and evaluation rely on data collection instruments to elicit
and record information.
O There are various types of instruments available including summary
report forms used to aggregate project data, questionnaires, more
informal interview guides, inventory forms, and observation checklists.
O Several formats can be combined on one form.
O For example, a household survey may include questions for the mother
about her familys health as well as a checklist for observations about
the size, construction, and cleanliness of the house.
O The development of clear, easy to read instruments is a vital
component of both monitoring and evaluation.
O Knowing what to include and how to organize the questions and
answers requires skill, common sense, and practice.
O Considering how the collected data will be analyzed and used before
designing the data collection instrument, creates a focused, efficient
instrument. outlines considerations to be taken when designing data
collection instruments.
O Clear language and an easy to follow format are essential to collect
accurate, useful information.
Steps in preparation for DCI:
1) Draft the content of the instrument based on pre-determined
information needs.
Information should be collected and recorded in the simplest way
possible while maintaining reliability.
Those designing the data collection system should determine in
advance how each piece of information collected will be used.
It is useful to keep in mind that the collection, processing, and
analysis of information is costly in terms of financial and human
resources.
Keep it short and simple.
Identify subject and sources.
Use filter questions
Calculate later
The instrument must provide researchers with an easy way to
identify the child, household, clinic, or community being studied.
Codes are often used to identify the physical location of the
subject (e.g. a province, state, district, county, village, cluster,
and house) and the particular unit of analysis (e.g. a clinic, a
household, an individual).
In addition, the persons filling out the forms must be identified
on the form to avoid asking inappropriate questions. For
example, ask a woman if she has children before asking the age
of her youngest child.
2) Pay attention to language.
When designing monitoring and evaluation instruments, it is
important to use the local spoken language and in ways which
capture local concepts.
Sometimes pictures or symbols that are easily recognized by
respondents are useful to convey ideas, particularly where level
of literacy is low.
When discussing past events or seeking to estimate the age of a
person, it may be helpful to use local holidays, seasons (e.g., the
harvest or the rainy season), and special occurrences (e.g., a
flood or construction of a school) to help respondents with recall.
When working in foreign languages, a good translation is
essential.
If the original instrument is to be written in English, it is useful to
translate the English version into the local language and then
have someone else translate it back into English.
Not only will this prevent translation mistakes, but it will also
allow investigation;
Field Insight: Defining Terms to Avoid Misinterpretation
Sometimes what is clear to those designing instruments may be
incomprehensible or misinterpreted by others.
This is particularly true when translating instruments into foreign
languages. Even in the same language, however, there may be
different understandings of concepts. In Ghana, for example, the
term short-term hunger, which was introduced by a workshop
facilitator to discuss the consequences of skipping one or two
meals, was interpreted by workshop participants as the seasonal
hunger that local people experience six months of the year.
Avoid using double negatives such as, Do you believe that a
mother should not have to attend feeding sessions if her child is
not at risk?
Be specific. Language used, whether in interviews or on reporting
forms, must be specific to prevent confusion. Instruments should
avoid general terms such as small, big, frequently, or
often.
Employing the following strategies when crafting questions can
prevent this phenomenon. That present two ideas
simultaneously such as, Do you think children should receive
deworming medicine and micronutrient supplements?
Make sure that each question covers a single idea and refrain
from using words that could influence answers. For example, the
question, Do you think it is healthy to take a daily vitamin
supplement is a slanted question, because it presents the idea
of healthy without the opposite concept.
A better question would be Do you think taking a daily vitamin
supplement is good for your health, bad for your health, or
makes no difference? While this may sound cumbersome, it will
assure that the use of a positive word in the question doesnt
bias the answer.
Another way of asking the question would be What do you think
about taking a daily vitamin supplement? and recording the
answer in pre-coded categories.
3) Craft questions carefully.
Carefully craft questions.
Inappropriate wording of questions can influence responses and
may lead to information which do not accurately reflect the
situation.
Employing the following strategies when crafting questions can
prevent this phenomenon.
Avoid double barreled questions that present two ideas
simultaneously such as,
Do you think children should receive deworming medicine and
micronutrient supplements?
Make sure that each question covers a single idea.
Keep questions neutral and refrain from using words that could
influence answers.
For example, the question, Do you think it is healthy to take a
daily vitamin supplement is a slanted question, because it
presents the idea of healthy without the opposite concept.
A better question would be Do you think taking a daily vitamin
supplement is good for your health, bad for your health, or
makes no difference?
Do not reference authorities. Questions can also be slanted by
using references to authorities:
Do you agree or disagree with the nurses advice to breast-feed
exclusively for four months? appears balanced, but reference to
the nurse suggests that one answer is better than the other.
Respondents may give an answer which reflects what they think
is expected rather than what they truly believe.
Avoid hypothetical questions. All questions should refer to actual
events and circumstances.
Hypothetical questions frequently are difficult to conceptualize,
and often generate confusion.
Ask for facts, not judgments. Ask for facts or descriptions rather
than judgments on the adequacy of the service.
For example, ask, How many meals did each household member
consume each day during the past week? rather than Is your
family short of food?
The definition of short of food may vary between participants
and therefore, be difficult to interpret during analysis of the data.
Subjective questions can be useful, as long as it is recognized
that they measure opinions and attitudes, not facts.
4) Optimize the sequence of the questions.
Begin with a non-threatening question.
Always begin an instrument with easy, non-threatening questions
which will build rapport with the respondent.
Questions should be grouped by topic and generally within each
topic by response type (e.g., yes/no, agree/disagree, and multiple
choice).
Place open-ended questions and questions about sensitive topics,
such as income, near the end. Some surveys end with an easy to
answer question; in an effort to leave participants feeling positive
about responding.
5) Make the format easy to follow.
Pay attention to layout. The format of a data collection
instrument, whether for a one-time survey or for ongoing
monitoring, is often as important as the content itself. A poorly
laid out instrument will result in errors, and information will be
unreliable. It is important to keep the instrument uncluttered and
with plenty of white space between questions.
Questions should be clearly numbered and ample room provided
for answers. Boxes or lines should be used to contain responses.
Each form should be clearly labeled at the top with both the
name of the form and the level of data collection (e.g., health
clinic, district office, etc.) with a space to indicate the date of
data collection. If the instrument is longer than one page, the
pages should be stapled together with identification