Home > MB313 > Quantitative Proposal
MB 313
PROPOSAL GUIDELINES
QUANTITATIVE PROJECT
The
proposal should be a maximum of two double spaced pages covering the following
points:
RESEARCH
OBJECTIVE
What
do you hope to accomplish with this research?
What is your research question?
HYPOTHESES AND EXPECTATIONS
What do you expect? What are your independent and dependent variables? Draw diagrams of how you expect your independent and dependent variables to be related in your test of hypotheses. Your hypotheses and expectations will give you guidelines for what data you want to collect.
METHODOLOGY
How
are you going to conduct your research? You need to talk about:
Technique – You will be doing survey
research. Attach a draft of your questionnaire so we can discuss it. The questionnaire
should be at least two pages long. The questionnaire should also contain a scale. You can make up the
scale or get a scale from previous studies. The questionnaire should also contain some demographic information
such as sex and/or class of student. You need to pay attention to transitional
statements, skip patterns, question wording, question grouping, and question
order. Here is a cover letter to go with the
questionnaire
Sampling – how are you going to choose
subjects? I think that about 30 respondents is a large enough sample for this
project.
Pretest – You will need to pretest
your questionnaire on at least 3 people. Make notes on what is said and how
it led you to change your questionnaire.
For example, did the respondents understand your scale? You do not have to pretest before meeting with
me, but it is a good idea.
Data
Coding
– what are your plans for coding or describing the data? You will need to create a codebook.
To be more specific
Creating a Scale:
You will need to create a scale as part of this project. A scale is composed
of a number of statements about something that you want to study. You should
have at least four items in your scale, but more is better. The items should
be expressed as declarative statements that the respondent can have an opinion
about. For example, if you wanted to study whether students were polite, you
could create a scale of politeness based on the students opinions about whether
they engage in rude behavior that could look something like this:
1. You should send a written thank you note if you receive a present.
2. You should rarely honk your horn while driving a car.
3. You should say "bless you" if someone sneezes.
4. You should hold doors open for others who are behind you.
For this scale you could create response categories such as "strongly
agree", "agree", "neutral", "disagree",
or "strongly disagree". Then you could code the responses from 1
to 5, add up the responses for a respondent, and get a attitude towards politeness
scale value. Then you might want to test hypotheses such as males score higher
on politeness than females. Your independent variable would be gender and
your dependent variable would be the politeness scale. You might also have
a continuous or interval level scale variable that would predict politeness,
such as the number of hours of sleep you get a night. In this case you would
predict that number of hours of sleep would predict your politeness score.
There are several points to keep in mind when creating a scale. First, you
have to come up with statements that the respondent can agree or disagree
with. If you asked "How polite are you?" or "Do you think you
are a polite person?" then people will want to answer yes or no and a
5 point agreement scale would not make much sense. This is why you need to
express your scale questions as declarative statements such as "The world
is reaching the limits of people it can hold", so that people can have
levels of agreement or disagreement. Another point to keep in mind is that
all of your statements have to address the same underlying topic. If you added
a statement to your politeness scale such as "I always yell at football
games", you may not be measuring politeness, but rather excitement at
sports contests. It is a good strategy to think of more statements than you
are going to use in your final scale and then give the scale to some people.
See if they consistently agree or disagree with the items, and, if they are
not consistent in their answers, ask them why they scored some items as agree
and some as disagree. You may find that they did not interpret all of the
items as measuring what you thought the items would measure. In this case,
you can delete some items and develop a more accurate scale. Finally, don't
hesitate to look on the internet or the research literature for scales. You
should feel free to use published scales in your project. You can search our
library's databases
of psycINFO or psycARTICLES of academic psychology articles using psychological
scales. Occasionally the article will list the items used in a scale in an
appendix. You can also go to the Research Methods Division of the Academy
of Management and click on "Measure
Chest" to find a database of scales and references of where to find
them.
Example: A previous group studied cell phone use among the undergraduates.
Some of their initial hypotheses were that women would use cell phones
more than men and type of phone plan would affect usage. They could have compared
number of minutes used per week by men and women by by looking at two boxplots.
This would tell them the difference between men and women (categorical or
nominal variable) on number of minutes used (continuous or interval variable).
The difference could be tested with a t-test. They could have created a scale
with a number of questions about how convenient the respondent found cell
phones. By adding up the responses
to these attitude questions, the group would create a continuous or interval
variable measuring satisfaction with the cell phone.
They
could have proposed a causal hypothesis: higher satisfaction with their cell
phone leads to more cell phone minutes.
Then
they could have done a regression test using satisfaction (independent variable)
to predict minutes of cell phone use (dependent variable).
Other examples of student projects:
Research Question: Yankees Fans vs. Red Sox Fans: Who is Smarter?
Hypotheses: Red Sox fans are more enthusiastic about their team than Yankees
fans. Yankees fans are more academically dedicated than Red Sox fans. The
more enthusiastic fans are about their baseball teams, the less academically
dedicated they will be.
For this study the students had to generate scales for baseball team enthusiasm
and academic dedication. For example, academic dedication was measured by
"How many books did you read during the summer of 2xxx?", On average,
how many hours each week do you study for tests?", "On average,
how many times each month do you skip class?". Note that in this case,
the students are created a scale based on how many times certain behaviors
take place. They had the respondents checking off categories such as "0",
"1-2", 3-4". Then they coded the categories as 1, 2, 3, etc.
and added them up to get a score. I am not sure that the number of books read
during the summer measures "academic dedication", and I would have
suggested dropping this question and focusing on academic school year efforts.
These questions could also be focused by asking "Last month, how many
times did you skip class?" rather than "On average ...". Recall
of behaviors is likely to be more accurate if the respondent is thinking of
more recent behavior.
Home
> MB313 > Quantitative Proposal
This page
last updated
11/01/2006