Chapter 4: Research Plan
(F) Day of the week: Tuesday Class: IS303 Created Time: November 10, 2020 7:38 AM Database: Class Notes Database Date: November 10, 2020 7:38 AM Days Till Date: Passed Last Edited Time: June 9, 2021 10:42 AM Tags: Research Plan Type: Reading Notes
Content
Every research project shows the same content headers
- Measurement
- Sampling
- Data collection Techniques
- Research Plan etc.
1. The Research Plan
1.1. Coming up with the Right Question
Research Questions must be
- Researchable: the question must be answerable
- Interesting: to be ‘green-lit’ by funders, the research must attract attention
- Resolving conflictual research debate topic
- Extend research questions more inadept
- Using new methodology
- Address a knowledge gap
1.2. Reviewing the Literature
Review past relevant research literature on the topic being researched
- to see how past-researchers approached the topic
- to see how effective their research questions were
- to guide on vocab usage, data collection method…
To start researching a new topic you must catch up on past findings.
You must narrowly scope your questions to be able to consume all the past research available.
On a broad subjects
- You read titles and abstracts to get the basics
- Then insert specific term to narrow the scope
1.3. The Case and Place for Theory
-
The question must have theoretical connection to the topic to be interesting
|Create a theory that can be tested against Empirical evidence
-
The question must have a purpose for being asked, and how it contributes to the research
-
Topic must not be abstract as to not able to be tested by empirical evidence
Therefore
The “middle-range theory” is the goal for a research to be at.
|For empirical evidence to repeatedly confirm the theoretical ideas.
2. The Placement of Theory in the Research Process
2.1. Deductive Research
where data collection is the starting point to forming a theory
Empirical generalizations – generalizations based on or drawn from observed patterns in the data.
Look for patterns in the collected empirical evidence
- Strengthen what is already known from empirical generalization
- Then generate a corresponding theoretical idea
2.2. Inductive Research
Definition: where the theory is used as a start for data collection
Propositions or hypotheses derived from the theory to be tested with empirical data.
The test will either:
- support/reject the hypothesis
- strengthen/challenge the original theory
3. The Levels of Theory Guiding or Resulting from Researchhere the theory is used as a start for data collecti
3.1. Macro Theory
Are theories having a broader scale in explaining social processes or institutions
Ex: Macro-economic studies an entire country’s:
- Unemployment
- Trends
- Productivity
3.2. Micro Theories
Are theories focusing on narrower scope involving smaller slices of social reality
Ex: Micro-economics focus on behavior of specific individuals or businesses
3.3. Meso-level Theory
It tries to merge macro- and micro-level in an intermediate level of theoretical explanation
It focuses more on groups or formal organizations
3.4. Paradigms
Paradigms are framework of thought that gives very broad views of how the world works. It makes assumptions and methods for understanding the world.
3.4.1. Order Paradigm (Macro)
The world is consensus-based orderly arrangement of social institutions that compliment each other, making a smooth social system.
Concepts: social solidarity, social stability, social integration, collective conscience
3.4.2. Conflict Paradigm (Macro)
The world is a non-consensus place and any sense of order comes from power and coercion.
Concepts: coercion, power, force, special interests
3.4.3. Symbolic-interactionist Paradigm (Micro)
The order around us are result of interactions and negotiations between people in society
Concepts: impression management, social context, verbal and non-verbal communication
4. Causal vs. Non-Causal Question
Causal Question
Is to ask if something is the cause of something else.
The experimental design: to manipulate independent variables and observe for changes
- Experimental Group: exposed to the independent variable
- Control Group: not exposed to the independent variable
- Random Assignment 1: equal chance to end up in control group or experimental group.
Non-Causal Question
Non-Causal Questions uses Cross sectional design: ‘What’ questions that pursue
- description/Explorative
- correlational analysis
- exploration or evaluation
- process or change
Non-experimental research designs: (survey)
- Cross-sectional design
- Longitudinal Design
- Examples
- Survey description of population
- Social Process: sex-change, terrorist-recruitment
- Impact of social programs/policies
- Social Change
5. Timing
Research might want data to be in a specific moment in time or to be for change over time.
5.1. Cross-Sectional Research Design
Is to collect data at a moment in time
Data collection happens through surveys
Used for:
- Social Research: people’s behavior and thought
- Correlational Analysis: associations between variables in society or social process
5.2. Longitudinal Research Design
Analysis of change over time. It collects data at 2 or more points in time. (Wave of study)
Change happens at 2 basic levels:
- Individual level: requires fixed sample panel design
- “Higher” group or social level: repeated cross-sectional design of differing samples
Fixed-Sample Panel Design
Used look for change in the individual level of analysis
Data (survey) is collected from the same sample of respondents at multiple points in time.
It’s difficult to execute, why?
- Task of Tracking People: people inform their friends and family about their whereabouts, but not researchers.
- Panel Mortality: subjects drop participation half way through
Repeated Cross Sectional Design
data (survey) collected at two or more points in time from different samples of the same research population each time
It can measure change in population, but not individuals overtime
Cautions
- Fair sample the representation of the same population on each wave
- Consistency in the measurement process, same survey questions…
Event-Based Design
Data (survey) collected of change in cohorts who share a defining event.
Cohort: people defined by the same event. (Boomers)
6. Field Research
Research about processes requires qualitative research plans for Fieldwork
- Field Workers: are Researchers meet subjects and gather details about their process
- Intensive/unstructured interview: is hands on research in the subject’s natural habitat
7. Validity
7.1. Measurement Validity
Selection of key variables that most relevant to the research topic
- Concepts could be turned into measurable variables
- Out of the possible measures in the project, pick ones that most accurately define causation in the research
- Literature Review past variable choices and their effectiveness, could help with choosing variables
7.2. Internal Validity
To select the appropriate strategy of action to the research agenda
Ex: Research of Causal Analysis must use Experimental designs are the best for isolating the cause and effect relationship and thus are strongest for achieving internal validity.
7.3. External Validity
Ensure the generalizability of the research findings by carefulness in sampling procedures
How?
-
Probability Sampling: individuals are selected by RNG without biases from the population.
-
Experimental Design: fails to replicate real life, out of lab experiences (mundane realism)
⇒ to gain external validity the experiment must be repeated with different conditions to see if the finding holds
Conclusion
- A theory will direct research in selecting concepts, variables, and hypotheses or questions.
- Specifying research question to decide between experimental or non-experimental design
- Plan for time specific or change over time data collection, longitudinal or cross sectional
- Validity of measurement, internal, and external brings trust in the research
3 Main Research Desing
- Experimental Design
- Cross-sectional Design
- Field Research
Content
https://drive.google.com/file/d/1qvW13G_puo4QNbhltwLfLg0-RMLci8FZ/preview
Footnotes
-
The picking of random samples from a group without discrimination, bias, or favor towards any characteristics. ↩