Please look at this tutorial, if you are going to be using regression analysis in your paper:
http://academic.udayton.edu/gregelvers/psy216/spss/reg.htm
http://academic.udayton.edu/gregelvers/psy216/spss/reg.htm
29 Aralık 2015 Salı
23 Aralık 2015 Çarşamba
Data Categories
• Data categories should
be mutually exclusive and exhaustive.
•Mutually exclusive: each possible survey
response should belong to only one category, whereas
•Exhaustive: categories should
cover the entire set of possibilities.
Data Categories
•Nominal
data: different categories of
data, but they lack an order. (A binary variable is a nominal variable with
only two categories.)
Examples: sex (binary),
name, nationality, eye color, city
•Ordinal
data: Categorical data that
has an intrinsic order. It only shows sequence. The distance between categories
is not really standart, so we cannot add, subtract, etc.
Examples: Likert items
(level of agreement), restaurant/hotel rankings, race rankings (first and
second in a competition).
•Discrete
data: only has integers and
decimals are meaningless. We usually collact discrete data by counting
items/events.
Example: The number
of children in a family. The number of votes for a party.
•Continuous
data: can take any range of
values.
Examples: weight,
height, age, heat.
•Interval
data: Each position is equally
distant from one another, but ratios are meaningless. Zero is not meaningful
(it is arbitrary)
Example: Heat. 20 degree
and 25 degrees are 5 degrees different. 35 degrees and 40 degrees are also 5
degrees different. But 40 degrees is NOT twice as hot as 20.
•Ratio
Data: Numbers can be multiples
of each other. can be multiplied and divided. Zero has a meaning.
•Example: Age, weight,
height.
QUALITATIVE METHODS
There are ways of
conducting qualitative research. We will cover 2 of the most common methods:
1.
FOCUS GROUPS:
Morgan,
D. (1996) Focus Groups. Annual Review of Sociology. 22: 129-152.
“Collects
data through group interaction determined by the researcher.” P. 130
It is
INTERACTIVE.
Generally
4-10 people. Smaller groups are preferred for emotionally charged topics. They
are easier to mdoerate. They allow more sharing. There is less idea generation.
They can
be strangers, or acquaintances. (what you prefer depends on your research
question).
Goals:
İdea generation
Understanding what people think and how people
talk on a topic
Possible problems:
Conformity
The group effect: people query each other and
explain themselves to each other. We can see the extent of consensus and
diversity of the participants.
The role of the moderator is important: do not
interrupt the group interaction.
Some topics may be unacceptable for group
discussion.
Questions: should the questions be standardized?
How structured should the interview be?
Sampling: segmentation – consciously varying the
composition of the groups (socioeconomic group, geography, education, sex,
age).
1.
Adds a
comparative dimension
2.
Facilitates
discussion by making discussants similar to each other.
Saturation: after a few groups, the discussion
starts repeating and the researcher can predict what they participants will
say.
Time:
Average is 90 minutes.
2.
IN-DEPTH
INTERVIEWS:
Be fully
prepared.
Cover
all publicly available resources on the topic (Literature review).
Targeted
and efficient questions.
Develop
hypotheses in advance about what you expect.
Be open
to have your hypothesis falsified
May
conduct exploratory interviews with low-ranking people
Start
interviews with low-ranking people
Record the
interview IF POSSIBLE. Transparency!
HOW TO DO SCIENTIFIC RESEARCH
1.
Choose a
research topic
2.
Formulate a
research question and testable
hypotheses
3.
Conduct
literature review
4.
Define the
concepts
5.
Decide how to
measure your variables (Operationalization)
6.
Collect data
using qualitative and quantitative methods.
7.
Analyze and
discuss your findings.
OPERATIONALIZATION
Look at the literature on
the topic:
How did
others opertaionalize it before you?
Are
there ready-made scales?
Is there
a ready data set that you can use?
If you need to create your own data set, what type of data do you need? The hypothesis should determine what type of data is necessary.
1.
Qualitative
methods:
collect non-statistical findings
not generalizable: findings are not representative of a population.
deep explanations: you get a very deep understanding of a small group of people’s perspectives, or a few cases.
collect non-statistical findings
not generalizable: findings are not representative of a population.
deep explanations: you get a very deep understanding of a small group of people’s perspectives, or a few cases.
When you are reporting results, you
should not generalize.
Qualitative data accumulates and gives
us very valuable information.
It can be used in conjunction with
quantitative data.
2.
Quantitative
methods:
statistical data,
representative: the data represents a
broader population (a school, a city, a nation, etc.) . It can be generalized
to that population.
broad explanations: you get a
superficial understanding about the whole universe.
MEASURING THE RIGHT THING:
1.
VALIDITY
Am I
measuring the RIGHT THING? Am I measuring what I am supposed to be measuring?
Or am I measuring something else?
Measuring
democracy:
We can
ask people, if they like democracy. Let
us say that 90% of the people in a country say that they like democracy. Does
that mean that country is democratic? No. We are measuring the wrong thing.
2.
RELIABILITY
Am I
measuring it CORRECTLY?
Measuring
Democracy
To
measure democracy, we look at the number of incarcarated journalists.
The government says that there are no
incarcarated journalists. However, Journalists Without Borders say that there
are 248 journalists in prison. Which data is more reliable? Which one are we
going to use?
Measuring alcoholism:
Ask people how much they drink per week.
Reliability problem: people lie about how much
they drink. OR They may not know how much they drink.
Literature Review
Literature Review
1.
Write a literature review on
any topic of your choice.
2.
Your review should be a minimum
of 200 words and maximum of 300 words.
3.
It should include a minimum 5
and maximum of 8 social science journal articles.
4.
No books, websites, news
articles or other sources will be accepted. You are ONLY allowed to refer to
social science journal articles.
5.
ALL of your sources should be
in English. Sources in Turkish are not allowed.
6.
Read the literature review
samples I handed out in class. Those are GOOD examples. You can find more
examples in political science, sociology and psychology journals.
This is an excellent source on how to write a literature review.
Read it very carefully before you write your literature review.
Also, have a
look at this link:
And this:
Online Research
ONLINE RESEARCH:
There are many
ways of conducting online research. Here are two common and practical methods:
METHOD 1:
STEPS
1.
Download Zotero as an extension
or stand alone:
2.
Go to the databases on the IKBU
library website:
http://www.kemerburgaz.edu.tr/egitim-hayati/liste.aspx?SectionID=kAR%2fUExqe98m13gkIePW2w%3d%3d
http://www.kemerburgaz.edu.tr/egitim-hayati/liste.aspx?SectionID=kAR%2fUExqe98m13gkIePW2w%3d%3d
3.
Pick a database: e.g. JSTOR
http://www.jstor.org/
http://www.jstor.org/
4.
Search one or more terms: e.g.
TERRORISM
5.
Download the article on Zotero,
if it is free: e.g.
Krieger,
T., & Meierrieks, D. (2011). What causes terrorism? Public Choice, 147(1/2), 3–27.
6.
Choose from the articles that
are cited in the article. Go to “References” section. Choose the ones that are
relevant to your research. Click and download.
7.
Prepare an annotated
bibliography.
METHOD 2:
STEPS:
2.
Search a term. E.g. terrorism
3.
Limit the dates. E.g. since
2000
4.
Choose an article.
5.
Click on cited by…
6.
Go to the page, where all the articles
that cited this article are listed:
7.
Find the articles/books that
are relevant to your research.
8.
You can reach some
articles/books by clicking the links on Google Scholar. If you cannot, go to
the IKBU library website.
9.
Search those articles on the
library website. E.g.
Pape, R. A. (2003). The Strategic Logic
of Suicide Terrorism. The American Political Science Review, 97(3),
343–361.
9 Aralık 2015 Çarşamba
Kaydol:
Kayıtlar (Atom)