Stages in the research process

Formulate Problem

Determine Research Design

Topic 1 | Introduction Design Data Collection

Method and Forms

Design Sample and Collect Data

Analyze and Interpret the Data

Prepare the Research Report

ESC5517 | Week 1 | 12 September 2015| BSAS 1 @ ENV

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 1

 Overview of research design Overview of research design
Exploratory Uses Types
•Formulate problems more precisely •Literature search
• “discovery” •Experience survey
•Develop Hypotheses
Descriptive Exploratory •Establish priorities for research •Analysis of select cases
Research •Eliminate impractical ideas •Interviews
• “relationships” •Clarify concepts •Ethnographies
Causal •Focus groups
•Etc.
•Describe segment characteristics
• “cause-and-effect” Descriptive •Estimate proportion of people
Research who behave in a certain way •Longitudinal study
•Make specific predictions •Panels
•Sample Survey
Causal •Provide evidence regarding causal
Research relationships
•Laboratory experiment
•Rule out all other explanations
•Field experiment

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 2

 Relationship among research designs Qualitative versus Quantitative research

Data
Quantitative = numeric data
Descriptive Research Qualitative = non-numeric data
• Caveat – all qualitative data can be coded and all quantitative
data is based on judgment
Common Assumption:
Exploratory Research
• Qualitative Data = preliminary
• Quantitative Data = confirmatory

Causal Research

Qualitative vs. Quantitative research Qualitative vs. Quantitative research

Qualitative Research Quantitative Research Detail versus Generalizability

To gain a qualitative To quantify the data and • Quantitative Methods

Objective
understanding of the generalize the results from the • Generalizations to other populations and/or situations
underlying reasons and sample to the population of
• Qualitative Methods
motivations interest
• Rich understanding
Sample Small number of non- Large number of
representative cases representative cases

Data Collection Unstructured Structured

Data Analysis Non-statistical Statistical

Outcome Develop an initial Recommend a final course of

understanding action
29

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 3

 Focus groups

Focus groups
• small group discussions led by a trained moderator

Objectives:
• Generate ideas
Exploratory research • Understand consumer vocabulary
• Reveal consumer needs, motives, perceptions, and attitudes
on products and services
• Understand findings from quantitative studies

32

Focus groups In-depth interviews (IDIs)

Advantages: In-Depth interview

• Generation of fresh ideas A set of probing questions posed one-on-one to a subject by a
trained interviewer so as to gain an idea of what the subject
• Client interaction
thinks about something or why he or she behaves a certain way
• Versatility
• Ability to tap special respondents

Disadvantages:
• Representative of the population?
• Subjective interpretation
• High cost-per-participant

33
34

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 4

 Popular qualitative research methods Still other qualitative methods

Ethnographies
• developing understandings of the everyday activities of people in local
settings
Focus Groups Interviews Observation
•Qualitative
•Flexible • Insight into actual, not reported, behaviors
•Probing
•Group dynamics, •Richness of data •Not influenced by others Protocol analysis
expect more creative •Gets at the “Why” •Max probing, great depth
of customers’ •Expensive
• involves placing a person in a decision making situation and asking him or
•Some probing
•Relatively inexpensive behaviors •Time consuming her to verbalize everything he or she considers when making a decision
•Ready industry •Generates ideas •Candid, sensitive topics
•Clarifies other Projective Techniques
project results • participants are placed in (projected into) simulated activities in the hopes
that they will divulge things about themselves that they might not reveal
under direct questioning

35

Descriptive research

For our purposes, “survey” research

• Asking a sample of people from a population a set of
questions
• Using the answers to describe that population
Common Goals
Descriptive research • Describe what is going on or exists
• Estimate how groups of consumers might behave
• Examine relationships between two or more variables
• Predict

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 5

 Example – examine the relationship Descriptive research

In contrast to exploratory (qualitative) techniques, the purpose

Differences between Estimated and Actual Calories of survey research is to produce statistics
In contrast to exploratory (qualitative) techniques, survey
Actual Calories Consumed research is concerned with generalizability
2000.00
• Sampling becomes very important

1000.00

ffi
d
l 0.00
a
c

-1000.00

-2000.00

0.00 500.00 1000.00 1500.00 2000.00 2500.00 3000.00

calactrc

Descriptive research Descriptive research

Three key aspects Two Basic Types

• Designing questions 1. Longitudinal
• Sampling 2. Cross-Sectional
• Data collection
We want to have a study that is precise, credible, and accurate
• A key issue is reducing error throughout the process
• Poor attention to ANY of these aspects can result in poor
results

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 6

 Limitations…

• Correlation is NOT causation

• Survey research is limited by what people are willing and able
to tell us in the context of a survey
• This limitation can be addressed using other methods
(triangulation)
Casual research

Causal research
• Helps us determine if one or more IVs (treatment, predictors)

causal
causes or affects one or more DVs (outcome variables)
• Most demanding design—strongest conclusion
• Requires the highest degree of understanding of the problem

- adjective

1. of, consisting, or implying a cause

2. Grammar. expressing a cause, as the conjunctions because and
since

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 7

 Evidence that supports a causal inference… Types of experiments
Laboratory Experiment
Correlation

• Observe the relationship (variability) between ad spend and Research investigation in which
investigator creates a situation
sales Experiment with exact conditions, so as to
Independent variable (treatment) occurs before the dependent control some, and manipulate other,
variables
(outcome) variable
Scientific investigation in which
• Change ad spend and look for an effect on sales. an investigator manipulates
and controls one or more Field Experiment
independent variables and
observes the dependent Research study in a realistic situation
variable for variation in which one or more independent
concomitant to the variables are manipulated by the
manipulation of the experimenter under as carefully
independent variables controlled conditions as the situation
will permit

48

Definitions and concepts Validity

Independent variables (IV) are variables or alternatives that are Internal validity refers to whether the manipulation of the
manipulated and whose effects are measured and compared, independent variables or treatments actually caused the
e.g., temperature observed effects on the dependent variables. Control of
Test units are individuals, organizations, or other entities whose extraneous variables is a necessary condition for establishing
response to the independent variables or treatments is being internal validity.
examined, e.g., sediment, water
External validity refers to whether the cause-and-effect
Dependent variables (DV) are the variables which measure the relationships found in the experiment can be generalized. To
effect of the independent variables on the test units, e.g., DO, what populations, settings, times, independent variables and
Extraneous variables are all variables other than the dependent variables can the results be projected?
independent variables that affect the response of the test units,
e.g., spatial, temporal

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 8

 Causal research (Experimental design) Causal research (Experimental design)

Internal Validity External Validity

Controlling extraneous variables Characteristics of “Good” experiments

Randomization Random assignment
• random assignment of treatment conditions to experimental Comparison group/control group
groups
• As a source of comparison
• Assumed to produce ‘balancing’ across groups
• As a control for rival hypotheses
• this is the key to internal validity (extraneous variables are
equal across groups due to random assignment) Generalizability/external validity
• Random selection

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 9

 Limitations of experimentation To conclude…
• Experiments can be time consuming, particularly if the Experiments are the only way to show causation
researcher is interested in measuring the long-term effects.
• But often take a back seat to descriptive studies due to time,
• Experiments are often expensive. The requirements of cost, and control issues
experimental group, control group, and multiple Exploratory and descriptive studies are useful, but be careful not
measurements significantly add to the cost of research.
to infer too much
• Experiments can be difficult to administer. It may be • Correlation is not causation
impossible to control for the effects of the extraneous
variables, particularly in a field environment. Again, let your research questions dictate your design!

• Competitors may deliberately contaminate the results of a

field experiment.

ENV @ UPM | @ZaharinAris | zaharin@upm.edu.my 10