framework for assignment on statistical analysis
TRANSCRIPT
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 1/8
MSc.Urban Ecological Planning (13-15)
AAR5250: PREPARATION FOR RESEARCH FIELD WORK METHODS
ASSIGNMENT
1. Please provid e a detailed answer to on e of the questions allott ed to you
usin g various references from th e literature and lecture notes. (minimum 2000
words)
2. How would you apply the learnings to your own research
proposal ?(minimum 1000 words)
Note: Question 1 has to be answered in group and question 2 have to be answered
individually by each member of the group. The names are mentioned against the
questions.
Question 4: What are the important considerations for designing a questionnaire in
research? How do we undertake statistical analysis of data gathered through
questionnaire?
RINY SHARMA
KARINA YUDONO
FIONA NSHEMERIRWE
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 2/8
Framework for assignment on Questionnaire design and Statistical Analysis
What is a questionnaire?
A questionnaire is defined as a formalized set of questions for obtaining information fromrespondents. The overriding objective is to translate the researcher’s information needs into
a set of specific questions that respondents are willing and able to answer. (Malhotra, 2004)
In a research, Primary data can be collected either through experiment or through survey
depending on the type of research whether qualitative or quantitative, Experimental or
Descriptive research. However, a questionnaire is the main means of collecting quantitative
primary data. The tool enables quantitative data to be collected in a standardized way so
that the data are internally consistent and coherent for analysis (Malhotra, 2004). Well-
designed questionnaires are highly structured to allow the same types of information to be
collected from a large number of people in the same way and for data to be analysed
quantitatively and systematically.
The questionnaire is mainly used in social science research. In this type of research, the
types of information will take account of the people’s or organizations’ level of knowledge,attitude, personalities, beliefs, or preferences. (Leung, 2001) The broad area of survey
research encompasses any measurement procedures that involve asking questions of
respondents. Quite often questionnaire is considered as the heart of a survey operation
(Kothari, 2004). Hence it should be very carefully constructed in order to obtain information
that is reliable, valid and truthful.
The use of questionnaires as a method of data collection is quite popular, particularly in case
of big enquiries. However, questionnaire is not mean to be the only source of information, so
the certain goals for doing questionnaire surveys has to be aligned with the features of
questionnaire.
Designing a questionnaire-
In designing a questionnaire it is important the following considerations must be made.
o The type of research whether Research question or hypotheses (how to test-
parameter and variable). The structure of the research questions in the
questionnaire is dependent on whether the research is exploring a given
phenomenon, or whether it is testing a hypothesis.
o If it is qualitative and quantitative data the researcher is interested in, this will
basically be reflected in the design of the questions.
It is important to pay attention to the sequence of questions in the questionnaire. The
first few questions are particularly important because they are likely to influence the
attitude of the respondent and in seeking his desired cooperation. The opening
questions should be such as to arouse human interest.
After the main objective of survey has been established, the detailed and
unavoidably tedious work of the questions shall design one by one. Questionnaire
have to refer on the data analysis plan, it will help to explain in detail what data are
needed to attain the objective set out for the questionnaire (Glewwe, 2005). To
facilitate a researcher, questionnaire can be built in several modules that aimed to
collect information on a particular subject.
o Depends on the objectives, question can be divided into open-ended and closed-
ended question. Open-ended questions allow the respondent to express an opinion
without being influenced by the researcher, while close-ended questions limit the
Formatted: Font: (Default) Arial
Formatted: Normal, Indent: Left: 0.28",bullets or numbering
Formatted: Indent: Left: 0.3", Hanging:0.2", Bulleted + Level: 1 + Aligned at: 0.78Indent at: 1.03"
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 3/8
respondent to the set of alternatives (Foddy, 1993). Open question can gather a lot of
information, however sometimes it difficult to summarize the result because of the
broadness. Closed question will make analysis easier but sometimes it failed to get
information aside from the given choice (Reja, 2003).
It is important to pay attention to the sequence of questions in the questionnaire. The
first few questions are particularly important because they are likely to influence the
attitude of the respondent and in seeking his desired cooperation. The opening
questions should be such as to arouse human interest.
Some people suggest that questionnaire should start with relatively easy to answer
insensitive question (Glewwe, 2005. Xxx,xxx). It believed can help interviewer build
ease. However, in practical lesson, there is one prove that sometimes hard question
should ask first because at the beginning of interview, respondent psychologically
still fresh to answer it. The easy question put in the end so tired respondent should
not think too hard to answer (xxx).
o Respondent/Target audience
Target audience will influence the questionnaire design because it effect
determination of the objectives from one to another questionnaire. It will be differ
according to what information need to be gathered from certain audience.
Different characteristic of the respondent should also taken into account when
designing a questionnaire so that respondent can give interviewer the needed
information.
o Limitation (time, budget, culture/habit)
There is low cost even when the universe is large and is widely spread
geographically using Internet based, Postal, and phone based questionnaires.
However this may not be the case in an interview based questionnaire that may
involve moving from door to door.
So many opinion about how to design a questionnaire and what is the better way to
do it in the field. Initial questionnaire is needed to make sure that questionnaire and
the procedure about how to do it is applicable (effective and efficient) in certain
condition. Means that it can reach the goals of the questionnaire survey.
Statistical Analysis
Types of data-(read more)
Variable can be divided by two, categorical variables and numerical variables.
Categorical variables are types or variables that are not numerical, for example
reason to take the bus. On the other hands, numerical variables are by their very
nature numbers (Glewwe, Paul & Levin, Michael., 2005). The main difference inthe treatment of those two types of variables arise from how to summarize it
when the catogories are large.
Categorical variables can be displayed by taking some part of variable or
aggregate it into broad categories. Numerical variables have several option of
displaying summary statistics.
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 4/8
Data entry- 1)Coding 2) Missing data 3) Filtering the data
Before doing the analysis, data should compile and manage. After entry data, it
should be checked at least twice to see whether there are errors or
inconsistencies. Weight need to be used in analysing survey daya and variences of survey
estimates need to be computed in a manner that reflects the complex sample
(Chromy, 2005)
Tools of analysis- 1) Frequency distribution 3) Central Tendency 3) Variability
The data collected can be thought of as a collection of variables, some of
variables are of interest in isolation and the others are interesting when compared
with the other variables.
If the number of values is big, it is good to show aggregated or summary
statistics concerning the distribution of the variables (Glewwe, Paul & Levin,
Michael., 2005).
Frequency Distribution:
Presenting all the numbers in each variables or categories is usually not as
helpful as showing the percentage of the observations.
Table 1.
Distribution of population by age and sex, Saipan, Commonwealth of the
Northern Mariana Island, April 2002: row percentage
Broad age group,
in years
Numbers Row percentages
Total Male Female Total Male Female
Total persons 67011 29668 37343 100 44.3 55.7
Less than 15 16915 8703 8212 100 51.5 48.515 to 29 18950 5765 13184 100 30.4 69.630 to 44 20803 9654 11149 100 46.4 53.645 to 59 8105 4458 3648 100 55.0 45.0
60 years or over 2239 1088 1150 100 48.6 51.4
Source: Glewwe, Paul., 2005. p. 338 (Round 10 of the Commonwealth of the
Northern Mariana Islands Current Labour-force Survey)
When presenting data in percentage, number can explain more. It is easier and
clearer to interpret, either to describe or compare to the others categories.
Yet, it depends on how to present the data, sometimes presenting numbers can
be work when putted in bar chart or line chart.
Formatted: Indent: Left: 0.54"
Formatted: Font: Bold
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 5/8
Figure 1.
Distribution of population by age and sex, Saipan, Commonwealth of theNorthern Mariana Island, April 2002: row percentage
Source: Table 1
Central Tendency:
The most common way to summarize numerical data is to show the average of
the data set. However, it can also eliminate unique event in data. For example in
income distribution, we can show the average if the gap between them is not too
big. If in one village, 1,000 people have income 50,000 NOK per month and 500
people have income 30,000 NOK per month. The mean will be 34,333 NOK.Nevertheless, the number is not presenting the situation since more people have
50,000 NOK for their monthly income.
To deal with such situation, median or midpoint of a distribution can help. In this
case, after arranging the entire number from the smallest to the biggest, the
midpoint is 50,000 NOK per month.
Figure 2.
Example for Median
0
5000
10000
15000
20000
25000
30000
35000
40000
Less than 15 15 to 29 30 to 44 45 to 59
Male
Female
Formatted: Font: (Default) Arial
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 6/8
To show the most frequently occurring value in data set, mode can be used. So
for the income distribution case, the most commonly occurring value is 50,000.
Variability:
Some commonly used measures the variability are range, midspread or inter-
quartile range, mean deviation, variance, standard deviation, and standard error
(Robson, Colin. (ed.), 2011).
Range can be used to show the difference between the highest and the lowest
score. In the example, the highest score is 50,000 NOK and the lowest score is
30,000 NOK.
Midspread is actually related to median. If median is 50 th percentile, than
midspread are the 25th and 75th percentile. In the given example, the 25 th
percentile is 50,000 NOK and 75th percentile is 30,000 NOK.
Mean deviation is the average of the deviations of individual scores from the
mean. The formula is |̅| . For the given example, the mean deviation
is 8,888.89.
Variance is the average of the squared deviations of individual scores from the
mean (Robson, Colin. (ed.), 2011). The formula is ∑ (∑ ) . In the
given example, the variance is 89,485,458.61.
Standard deviation is a square root of the variance. The formula is ∑ (∑ ) .For the given example, the standard deviation is 9,459.7.
Standard deviation shows distribution of the values, so it is better to put it in the
figure instead of number.
Standard error is the standard deviation of the mean score. Standard error is a
value measures how precisely the obtained mean value. The formula is √ .
0
10,000
20,000
30,000
40,000
50,000
60,000
Income Per MonthFormatted: Font: (Default) Arial
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 7/8
For the given formula, the standard error is 777.38 So, the mean is 34,333
±777.38 NOK
Summarizing/Inferencing from Statistical Analysis
8/12/2019 Framework for Assignment on Statistical Analysis
http://slidepdf.com/reader/full/framework-for-assignment-on-statistical-analysis 8/8
References
-How to design a questionnaire Wai-Ching Leung STUDENTBMJ VOLUME 9 JUNE 2001
studentbmj.com
-Research methods knowledge base, William M.K. Trochim, 10 October 2006
-Research Methodology –Methods and Techniques, C.R.Kothari, 2004, New Age
International (P) Ltd., Publishers, New Delhi
- QUESTIONNAIRE DESIGN AND SCALE DEVELOPMENT ,Naresh K. Malhotra, Georgia
Institute of Technology , Malhotra (2004)
Chromy, James R. and Abeyasekera, Savitri. 2005. Household Surveys in Developing and
Transition Countries. Chapter 9 Statistical Analysis of Survey Data. United Nation. New York
Glewwe, Paul. 2005. Household Surveys in Developing and Transition Countries. Chapter 3
An Overview of Questionnaire Design for Household Surveys in Developing Countries.
United Nation. New York.
Glewwe, Paul & Levin, Michael. 2005 Household Surveys in Developing and Transition
Countries. Chapter 16 Presenting Simple Descriptive Statistics from Household Survey
Data. United Nation. New York.
Reja, Ursa., Manfreda, Katja L., Hlebec, Valentina, and Vehovar, Vasja. 1993. Open-ended
vs. Close-ended Questions in Web Questionnaires.
Foddy, W. 1993. Constructing Questions for Interviews and Questionnaires: Theory and
Practice in Social Research. Cambridge University Press. Cambridge.
Munoz, Juan. 2005. Household Surveys in Developing and Transition Countries. Chapter 15
A guide for Data Management of Household Surveys. United Nation. New York
Formatted: English (U.S.)
Formatted: English (U.S.)