Practical Research: Planning and Design

Twelfth Edition, Global Edition

Chapter 6
Descriptive Research
Descriptive Research Designs
• Observation studies
• Correlational research
• Developmental design
• Survey research
Quantitative Observation Studies
• Involve humans, other animals, plants, nonliving objects
• Focus is limited, pre-specified
• Quantify behavior
• Require planning, attention to detail, and time
• Provide a quantitative alternative to qualitative approaches
Maintaining Objectivity in Observation
Studies (1 of 2)

• Define the behavior precisely and concretely

– Should be easily recognized
• Divide the observation period into small segments
– Record whether the behavior does or does not occur
Maintaining Objectivity in Observation
Studies (2 of 2)
• Use a rating scale to evaluate the behavior in terms of
specific dimensions
– Have people rate the same behavior independently
• Train the raters to use specific criteria until consistent
ratings are obtained
 Correlational Research (1 of 2)

• Examines the extent to which differences in one variable are

related to differences in other variables
• Researchers gather data about two or more characteristics for a
particular group to see if these characteristics are interrelated
 Correlational Research (2 of 2)

• Scatter plots show the overall pattern and describe the

interrelationship
• Correlation does not, in and of itself, indicate causation
 Developmental Designs

• Cross-sectional study
– people from several different groups are sampled and
compared
• Longitudinal study
– a single group of people is followed over time, and
data are collected at various times
Pros and Cons: Cross-Sectional V s. ersu

Longitudinal Study (1 of 2)

• Cross-sectional studies:
– Pro
▪ All the data can be collected at one time
– Con
▪ Different populations may represent different life
experiences (threat to internal validity)
Pros and Cons: Cross-Sectional V s. ersu

Longitudinal Study (2 of 2)
• Longitudinal studies:
– Pro
▪ Correlations between characteristics at different
times can be computed
– Con
▪ Participants may be lost to follow-up
▪ Characteristic being measured may change
because participants have experience with the
instrument
Compromise: Cohort-Sequential Design
• Addresses weaknesses of longitudinal and cross-sectional
designs
• Includes two or more age groups (the cross-sectional piece),
followed over a period of time (the longitudinal piece)
• Allows calculation of correlations between measures taken at
two different time periods
• Predictions can be made across time
Experience-Sampling Methods (E SM)
• An experience-sampling method (ESM) is an approach in which a
researcher collects frequent and ongoing data about people as they live their
normal, everyday lives
• Successfully used in both quantitative and qualitative projects
• Advantages of ESM methods:
– Potential for increased accuracy and validity of assessments
– Researcher gains data that might be useful in determining test-retest
reliability.
– Useful if the researcher wants to collect longitudinal data as a means of
investigating any short-term changes in characteristic as environmental
or behavioral variables change
Survey Research
• Goal is to learn about a large population by surveying a sample
of that population
• Also called a descriptive survey or normative survey
• Simple design – researcher poses a series of questions,
quantifies responses, and draws inferences about a population
• Captures a fleeting moment of time — extrapolation can be
made about a longer period of time
Types of Survey Research
• Interview
– Structured or semi-structured
– Face-to-face, telephone, video conference
– High response rate
• Questionnaire
– Paper-and-pencil or computerized
– Low return rate
– Assurance of remaining anonymous
Data Collection (1 of 2)
• Checklist: a list of behaviors, characteristics, or other entities
under investigation
• Limited information: observed or not observed
• Rating scale: used to evaluate a behavior, attitude etc. on a
continuum (“never” to “always”)
Data Collection (2 of 2)
• May be ordinal or interval scale
– People may not interpret scale the same way
• Rubric: two or more rating scales, with concrete
descriptions of behavior for each scale point
– Scales may not address the same things
Computerizing Observations
• Use a computer to record what you see
• Use a spreadsheet to organize the data
• Consider software specific to your purpose
Conducting Interviews in a Quantitative
Study (1 of 2)

1. Ask questions that help answer the research question

2. Write questions with quantifiable answers (numerical codes)
3. Restrict questions to a single idea
4. Consider asking a few questions to elicit qualitative data
5. Use a computer to streamline the process
Conducting Interviews in a Quantitative
Study (2 of 2)
6. Conduct pilot test(s)
7. Introduce yourself and explain the purpose of the study
8. Be sure participants offer informed consent in writing
9. Ask controversial questions in the latter part of the interview
10. Seek clarifying information as needed
Constructing a Questionnaire (1 of 3)

1. Keep it short
2. Keep the respondent’s task simple
3. Provide specific instructions
4. Use simple, clear, unambiguous language
Constructing a Questionnaire (2 of 3)
5. Give a rationale for any item for which the purpose is unclear
6. Check for unwarranted assumptions implicit in the question
7. Word your questions in ways that don’t give clues about
preferred or more desirable responses
8. Determine in advance how you will code the responses
Constructing a Questionnaire (3 of 3)
9. Check for consistency
10. Conduct one or more pilot tests to determine the validity of
your questionnaire
11. Scrutinize the almost-final product to make sure it addresses
your needs
12. Make the questionnaire attractive and professional looking
Using Technology When Administering
Questionnaires
1. Ask participants in the same location to answer directly on a laptop or tablet

2. When participants are at diverse locations, use email to request participation and
obtain responses

3. If you use paper mail delivery, us a word-processing program to personalize your

correspondence

4. Use a scanner to facilitate data tabulation

5. Use a computer database to keep track of who has responded and who has not
 Maximizing the Return Rate (1 of 2)
1. Consider the timing
– Avoid holidays and vacation times

2. Make a good first impression

3. Motivate potential respondents

– Write a great cover letter

– Include a self-addressed envelope with prepaid postage
– Offer to send the results of your study

4. If mailing your questionnaire, include a self-addressed envelope with return

package.
Maximizing the Return Rate (2 of 2)

5. Offer the results of your study

6. Be gently persistent
– Consider sending two follow-up reminders
– Send reminders a week or two after the previous mailing
Choosing a Sample in a Descriptive Study
• Any researcher who conducts a descriptive study wants to
determine the nature of how things are…
• It isn’t possible to survey an entire population of interest, so
researchers select a subset, or sample, of the population.
• A good sample is representative of the population.
• This is what quantitative researchers refer to as external
validity, of the extent to which the study’s findings can be
reasonably generalized beyond the partcipants in the study.
Probability Sampling
• Probability sampling: researcher specifies in advance that each
segment of the population is represented in the sample
– Requires random selection
Random Selection
• Each member of the population has an equal chance of being
selected
• Characteristics of the sample are assumed to approximate the
characteristics of the total population
• Tables of random numbers or computer programs are used to
select from a list of the population
Probability Sampling Techniques (1 of 4)
• Simple random sampling: Researcher numbers everyone in the
population and then uses random number generator to select
participants
• Stratified random sampling: Researcher identifies strata —
different groups in population — and samples equally from
each one
– Example: 10 students in each grade
Probability Sampling Techniques (2 of 4)
• Proportional stratified sampling: Researcher identifies strata
and samples from each one based on its proportion in the
population
– Example:
▪ population: 100 first graders, 200 second graders
▪ sample: 10 first graders, 20 second graders
Probability Sampling Techniques (3 of 4)
• Cluster sampling: Researcher subdivides a large area into
smaller units (clusters), selects a subset of clusters, and then
selects individuals randomly from each identified cluster
• Example:
– Population = all students in a district with 1200 schools
– Clusters = townships within the district
Probability Sampling Techniques (4 of 4)
• Systematic sampling: researcher selects individuals/clusters
according to predetermined sequence, which must originate by
chance
• Example:
– Scramble the list of people randomly
– Then pick every nth person
Nonprobability Sampling
• Nonprobability sampling: researcher cannot guarantee that
each element of the population will be represented in the
sample
– Some members of the population have little or no chance
of being sampled
Nonprobability Sampling Techniques (1 of 3)
• Convenience sampling (accidental sampling)
– Researcher takes samples that are readily available.
▪ Example: People who arrive at the store for breakfast
Nonprobability Sampling Techniques (2 of 3)
• Quota sampling:
– Researcher conveniently selects participants in the same
proportion that they are found in the general population,
but not in a random fashion
▪ Example population: 100 first graders, 200 second
graders
▪ Example sample: the first 10 first graders and the first
20 second graders who arrive at school that day
Nonprobability Sampling Techniques (3 of 3)
• Purposive sampling: Researcher choose participants for a
particular purpose
– People from voting districts that, in the past, have been
helpful in predicting the election outcome
– The researcher must always provide a rationale explaining
the selection of a particular sample
Surveys of Very Large Populations (1 of 2)
• Multistage sampling:
– Divide country into primary areas, randomly select areas to
sample
– Divide the primary areas into sample locations, randomly
select locations to sample
– Divide sample locations into chunks, randomly select
chunks to sample
Surveys of Very Large Populations (2 of 2)
• Multistage sampling:
– Divide chunks into segments, randomly select segments to
sample
– Divide segments into units, randomly select units to sample
Identifying a Sufficient Sample Size (1 of 2)
• Basic rule: The larger the sample, the better
• For smaller populations (N=100 or fewer), survey the entire
population
• If population is around 500, sample 50%
• If population is around 1,500, sample 20%
Identifying a Sufficient Sample Size (2 of 2)
• If population is over 5,000, a sample size of 400 is fine
• The larger the population, the smaller the percentage
• Need a representative sample
Sources of Bias (1 of 4)
• Bias
– any influence, condition, or set of conditions that
distort the data
• Researchers should try to avoid bias, but acknowledge
that it occurs
Sources of Bias (2 of 4)
• Sampling Bias
– occurs when any factor(s) leads to a non-
representative sample of the population
• Examples:
– Selecting from phone book (no land line?)
– Using an online survey (no Internet?)
– Mailing questionnaires (low or selective response
rate?)
Strategies for Identifying Sampling Bias
• Look for items that may be influenced by factors that
distinguish respondents from nonrespondents.
– interests, education level, age, etc.
• Compare responses that were returned quickly with those
that were returned later.
– late responses often look like what you’d expect from
nonrespondents
• Randomly select a small number of nonrespondents and
try to contact them.
 Sources of Bias (3 of 4)

• Instrumentation bias
– measurement instruments slant the results
– questions lead to particular answers
• Response bias
– participants say what they think researcher wants to
hear
– participants want to create favorable impression
(social desirability effect)
 Sources of Bias (4 of 4)

• Researcher bias
– researchers have a point of view
– researchers choose what they want to study
– researchers make subjective interpretations
 Interpreting the Data
• Don’t forget — it’s “descriptive” research, but you still have to
interpret the data
• Two basic principles of research:
1. The purpose of research is to seek the answer to a problem or
question in light of data that relate to the problem or question.
2. Although collecting data for study and organizing it for
inspection require care and precision, extracting meaning
from the data is paramount and should never be neglected.
 Questions to Ask Yourself (1 of 2)

1. Why is a description of this population and/or phenomenon

valuable?
2. What specific data will you need to address your research
problem and its subproblems?
3. What procedures should you use to obtain the needed
information? How can you best implement those procedures?
4. How can you get a sample that will reasonably represent the
overall population about which you are concerned?
Questions to Ask Yourself (2 of 2)

5. How can you control for possible bias in your collection of

data?
6. What will you do with the data once you have collected
them? How might you best organize and prepare them for
analysis?
7. Above all, in what ways might you reasonably interpret your
data? What conclusions might you reach from your
investigation?