Professional Documents
Culture Documents
AEB 303 Topic 5
AEB 303 Topic 5
How the data are collected depends on the research design process.
If survey research is used, then it can be collected through personal interviews,
telephone interviews, mail surveys, or online using a website or e-mail.
Surveys can be distributed at a mall, passed out in a class at a university, or stuffed in a
credit card bill.
The method that is chosen goes back to the three criteria already identified— time,
costs, and quality of information.
For observation research, multiple methods are available.
It can be done in person or with a video camera, Scanner data from cash registers can
be used.
Internet metrics can be utilized through cookie information for online observations.
Again, time, costs, and quality of information will impact the decision.
Methods involving human observation are almost always more expensive and take
longer.
Data collection can vary widely for experimental research.
It can take the form of questionnaires given at the end of an experiment that assess
attitudes, thoughts, feelings, or intentions.
It can be observation of human behavior, such as electronic tracking of eye movement
when viewing a print ad, or results of human actions.
With experiments, the type of experiment that is conducted will have the most impact
on how the data are collected.
Regardless of the method used to collect data, the goal should be to produce data that
are free of errors.
This requires identifying ways errors can possibly occur and then designing methods of
reducing the probability.
For instance, an interviewer’s facial expression or tone of voice in asking someone
questions may influence the respondent’s answers, creating error.
This type of error can be reduced by proper training of interviewers or by switching to a
self-administered questionnaire.
Errors can occur in recording of data or in selecting the sample.
While every study will have some random error, the goal of the researcher is to
minimize error as much as possible.
It is important to understand that in collecting data, errors should be minimized, and
steps should be taken in the design of the research to reduce errors.
Conducting a survey
A survey is a research method that gathers, examines, and analyzes information about
the opinions, characteristics, and behaviors of a group of people towards a particular
phenomenon, situation or topic of interest.
Surveys are conducted among predefined groups.
This means that a particular survey is conducted on a particular group of people who
have some characteristics in common.
Sources of errors
An error is the difference between obtained value and true values
It occurs when survey data (obtained data) do not accurately reflect the actual
behaviours, beliefs, and understandings of respondents in a population that a
researcher seek to understand
1. Errors in selecting the respondent
Sampling errors ( eg., using non-probability sampling method)
Coverage errors (e.g., a poor sampling frame omits certain groups of people)
Non-response errors at the level of a sampled unit (e.g. respondent refuse to
answer)
2. Errors in responding to survey questions
Non-response errors specific to a survey item (e.g. certain questions are skipped
or ignored)
Measurement errors caused by respondent (e.g. respondent does not listen
carefully)
Measurement errors caused by interviewers (e.g. interviewer is sloppy in reading
questions or recording answers)
3. Survey administration errors
Post-survey errors (e.g. mistakes in cleaning data or transferring data into an
electronic form)
Mode effects (e.g. differences due to survey method; by mail, in-person, over
the internet)
Comparability errors (e.g. different survey organizations, nations, or time periods
yield different data for the same respondents on the same issues)
Open-ended questions are questions that create room for explanations and elaboration
as opposed to closed-ended questions which require direct yes/no or true/false
answers and have no room for explanations.
Ordinal scale questions require answers that have some intrinsic ordering, for instance,
you can strongly agree, agree, be neutral, disagree, or strongly disagree with a
statement, or you can have a "low income", "middle income", or "high income".
Opposed to this are nominal scales which have no intrinsic ordering methods but just
unique categories, Examples of nominal scales may include gender, sex, place of
residence, etc.
Interval scales have no true zero point, as their zero values or points do not represent
the absence of data. Examples include temperature at 0 degrees Celsius doesn't mean
the absence of temperature.
On the other hand, ratio scales have a true zero point and never fall below zero. For
instance, the weight or height of an object can never fall below zero and therefore is a
ratio scale.
Online surveys
Online survey or internet survey, is one of the most popular data-collection sources,
where a set of survey questions is sent out to a target sample and the members of this
sample can respond to the questions over the world wide web.
Respondents receive online surveys via various mediums such as email, embedded over
website, social media etc.
Organizations implement online surveys to use the internet in order to gain insights and
feedback about upcoming products or services, change in marketing strategies,
enhancement in current features etc.
With the progress made by the internet, more and more organizations depend on the
data received and analyzed from online surveys to make integral changes in their
functioning.
Faster medium to reach the target audience: In comparison to the other survey
techniques such as pen-and-paper surveys, online surveys are exceedingly fast in
acquiring feedback from respondents.
Real-time analysis: After collecting information, the analysis is an important step for
researchers to take prompt actions. With online survey software available, researchers
can analyze collected data in real-time from a central dashboard.
Cost-efficient: Due to the minimum required resources, online surveys are much
cheaper than the other survey methods such as telephonic or pen-and-paper surveys.
Minimum margin-of-error: Unlike the traditional survey methods, respondents directly
answer the surveys, without the involvement of a mediator. The margin-of-error in case
of online surveys is hence, very limited.
Saves researchers’ time: Online surveys are quick to configure and send and due to the
ease, they tend to save researchers’ time.
Respondent Cooperation Issues: Online users are usually bombarded with surveys,
offers, and many other internet antics. There are chances that an important online
survey can be overlooked or left unanswered.
Questionable data reliability: Online surveys are conducted in the absence of a
mediator or interviewer. Due to this absence, the feedback received, in many cases may
be considered questionable.
Limited access to certain sections of a population: There are sections of a target
population which may not have access to the world wide web. In many cases, feedback
from such individuals may prove to be effective for research. For example, if an NGO is
conducting research to learn about the vulnerable members of a community, it is highly
unlikely that they have access to online surveys which makes it difficult for researchers
to collect data.
Collecting qualitative data: through Focus group discussion and Key informant interviews
A focus group is a small-group discussion guided by a trained leader.
It is used to learn about opinions on a designated topic, and to guide future action.
The main difference is the group has a specific, focused discussion topic.
The group has a trained leader, or facilitator.
The group's composition and the group discussion are carefully planned to create a
nonthreatening environment in which people are free to talk openly.
Members are actively encouraged to express their opinions.
Because focus groups are structured and directed, but also expressive, they can yield a
lot of information in a relatively short time.
Focus groups help people learn more about group or community opinions and needs. In
this respect, they are similar to needs assessment surveys.
Responses in a focus group, on the other hand, are typically spoken, open-
ended, relatively broad, and qualitative. They have more depth, nuance, and variety.
Nonverbal communications and group interactions can also be observed. Focus groups
can therefore get closer to what people are really thinking and feeling, even though
their responses may be harder -- or impossible -- to score on a scale.
Before you begin
It is used to learn more about opinions on a designated topic, and then to guide future
action.
The group's composition and the group discussion should be carefully planned to create
a nonthreatening environment, so that participants feel free to talk openly and give
honest opinions.
Since participants are actively encouraged to not only express their own opinions, but
also respond to other members and questions posed by the leader, focus groups offer a
depth, nuance, and variety to the discussion that would not be available through
surveys.
Additionally, because focus groups are structured and directed, but also expressive, they
can yield a lot of information in a relatively short time.
In short, focus groups are a good way to gather in-depth information about a
community’s thoughts and opinions on a topic.
1. Participants
Carefully recruited
2. Environment
Comfortable
Circle seating
Tape recorded
3. Moderator
Systematic analysis
Verifiable procedures
Appropriate reporting
2. Find a recorder.
A small but important point, often neglected.
You want to make sure people's ideas don't get lost.
Someone should be writing down what is said, in the same way as taking minutes at a
meeting. Arrange for this in advance. (Alternatively, you can audio-record, with the
group's permission. This will take more time -- to transcribe the audio, and interpret the
transcription-- but you will have a more complete, accurate, and permanent record.)
3. Decide who should be invited.
Ideally, those invited should be a representative sample of those whose opinions you
are concerned about.
Suppose you're concerned about the opinions of public housing tenants. You would
then want to spread your invitations across the different public housing facilities in your
community -- not just the best, or the worst, or the most vocal.
4. Decide about incentives.
That is, should you offer an incentive for people to participate?
Maybe not. In that case, why should people come? What's in it for them?
Possibly people will come just because they want to help.
Or because they think they will meet other interesting people, or learn something, or
just have fun.
Maybe the novelty of the experience itself will be a motivator.
And maybe all these reasons are true. (Or at least people believe them.)
But maybe those reasons aren't enough, and some other incentive is called for.
Money is one; sometimes focus group members get paid, even a small amount. (Focus
group leaders may get paid, too.)
If you can afford this, consider it.
If you can't, then think about other possible incentives: food and drink (more than chips
and soda?); public recognition; something to take home; a later training opportunity.
What will do the job?
5. Decide on the meeting particulars.
What day?
What place?
What time?
How long?
How many groups?
6. Prepare your questions.
Go in prepared.
Write out in advance a list of topics and questions you want to ask. This will serve as
your guide.
Below are some examples of general questions. These apply largely to groups discussing
a current program or service, but they can be adjusted for planned programs, as well as
for groups dealing with other concerns. The precise language and order of presentation
will depend on your topic and group, but some of these questions may be adapted to
your own needs.
"What are some of your thoughts about what's going on now?"
"Would you say you are satisfied with the current situation, with the way things
are going on?"
(If so) "What are you satisfied about? Why is that?" (Or, "What's going well...?")
"Are there things you are dissatisfied with, that you would like to see changed?"
(Or, "What's not going well...?")
(If so) "What are they? Why is that? How should they change? What kinds of
things would you like to see happen?"
"How about this particular aspect (of the topic). What do you think about that?"
7. Recruit your members.
Call them up. Email them. Or find them.
Remember:
8. Double-check.