COMP494 exam notes v2 Research Methods 2026 new update everything you
need to know before your exams Athabasca University
,Four types of research based on objectives.......................................... 1 Student T-test (Sample Statistical Techniques) .................................... 9
Problem Key Sections of the Paper ..................................................... 1 Choosing a Statistical Test................................................................ 10
Five Big Words in Social Research. .....................................................1 Data Mining, Knowledge Discovery (KDD), Kind of Data ....................... 10
Types of Research Questions............................................................. 1 Simulation ..................................................................................... 11
Optimization: Numerical Analysis, Numerical Methods, and Operations
Time-based research designs ............................................................. 2
Research ....................................................................................... 11 Modeling, Mathematical
Relationships in Research - how two variables are connected. ................. 2
Variables ........................................................................................ 2 Model, Bayesian ........................................... 11 Ethics, Refereeing,
Hypothesis ..................................................................................... 2 Plagiarism ......................................................... 11
Types of Data ................................................................................. 2
Surveys ....................................................................................... 12
Unit of analysis ............................................................................... 2
Correlation ................................................................................... 12
Two research fallacies ...................................................................... 2
Research Design............................................................................ 12
Hourglass Structure of Research ........................................................2
Types of Designs ........................................................................... 12
Components of a Causal Study ......................................................... 2 Research Methods in Information Sciences, Computational complexity
Positivism vs. Post-Positivism............................................................. 2 theory, Complexity and big-O notation.............................................. 13 Content Analysis,
Concept Mapping. ............................................................................2
Principal component analysis .................................. 14
Evaluation research ......................................................................... 3
Discourse Analysis, Social network, Social computing, Case Study
The Planning-Evaluation Cycle ........................................................... 3
Approach...................................................................................... 14
Study Questions .............................................................................. 3 Longitudinal Study......................................................................... 14
Types of Research Questions............................................................. 3
Action Research. ........................................................................... 15
Statistical significance ....................................................................... 3
Project Management.......................................................................15
Research Question and Hypothesis .....................................................4 Information Technology Infrastructure Library (ITIL) ........................... 15
Meta-Analysis ................................................................................. 4
Presentation, Research poster, Oral presentation, PowerPoint ................15
Sampling. ...................................................................................... 4
How to Write Research Paper, Thesis, Dissertation, References .............. 16
Probability sampling ........................................................................ 4 Qualitative Research, Measures, Debate, Data, Approaches, Methods,
Nonprobability sampling.................................................................... 4 Validity ........................................................................................ 16
Measurement – Construct Validity ......................................................4 Assignment 1 ...................................... Error! Bookmark not defined.
Types of measurement validity........................................................... 5 Assignment 2 ...................................... Error! Bookmark not defined.
The Four Levels of Measurement........................................................ 5 Four types of research based on objectives:
Descriptive – systematically describes a situation or phenomenon.
Software Quality ............................................................................. 5 Correlational – identifies relationships between variables.
Explanatory – explains why and how relationships exist.
Wechsler Adult Intelligence Scale ....................................................... 5
Exploratory – investigates little-known areas or research possibilities.
Modern Language Aptitude Test ........................................................ 5 Quantitative research (structured, preplanned, measures extent). Qualitative research
(unstructured, flexible, explores nature).
General issues in scaling .................................................................. 5
Thurstone, Likert, Gutman Scaling. ....................................................5 Problem Key Sections of the Paper:
Abstract: A brief summary of the entire study
Reliability ....................................................................................... 6 Introduction (includes Literature Review though untitled): research problem, relevant past
Validity ......................................................................................... 6 studies, rationale, research question and hypothesis
Method: Sample who participated in the study and how they were selected. It typically
True Score Theory ........................................................................... 7 includes: (sample size). Demographic details Sampling method, e.g.: Inclusion or exclusion
criteria, Measures instruments or tools used to collect data, Surveys, questionnaires, tests, or
Measurement Error .......................................................................... 7 observational checklists Descriptions of scales (e.g., Likert scales) and e.g. items. Design
Analysis ......................................................................................... 7 overall structure of the study, including: quantitative or qualitative, experimental or
observational, cross-sectional or longitudinal. independent and dependent variables, and how
Data Preparation (Analysis) .............................................................. 7 groups were assigned (randomly or not). Procedures step-by-step how the research was
conducted, such as: How participants were contacted and recruited. The sequence of
Descriptive Statistics (Analysis)......................................................... 7 activities (e.g., initial briefing, consent, data collection, debriefing).Duration Settings for data
collection, ethical considerations taken, like consent forms Results: Presents the data and
Inferential Statistics (Analysis) ..........................................................8
findings, statistical analysis, charts, or tables. Conclusions, References
T-test, Dummy Variables, GLM, Posttest-Only Analysis, Factorial Design Analysis,
Randomized Block Analysis, Analysis of Covariance, NEGD, RD
Five Big Words in Social Research
Analysis, Regression Point Analysis (Inferential statistical techniques)...... 8
Theoretical - ideas, models, or frameworks that explain how the world works. often seeks to
Event Data ..................................................................................... 9 develop, test, or refine theories.
ANOVA, Bayesian, Correlation, Factor Analysis MANOVA, Graph-Theoretic Empirical - observe and measure in the real w orld. collects data from experience, such as
Equation Modeling, Pearson Chi-squared Test, Regression Analysis, surveys, interviews, or observations. Nomoth etic - general case or universal laws rather than
,individual uniqueness. aims to find patterns that apply to groups or populations, not just one wording, cultural context). Qualitative data can be converted into quantitative form, such as
person. Contrasts with idiographic, which is concerned with the individual case. coding statements into numerical categories or similarity matrices. Eg grouping statements
Probabilistic - In the post-positivist view, research no longer claims to provide and representing similarities with a 0– 1 matrix.
certainty.Conclusions are probabilities, not absolute truths. findings are statistical estimates Unit of analysis - primary entity being studied or analyzed in a research project. what or who
with some (confidence intervals, significance levels). is being analyzed—not simply sampled—depends on how the data is analyzed, not just
Causal - cause-and-effect relationships—ultimate goal in much research is to determine how collected.
changes (interventions, programs, treatments) lead to changes in outcomes. Individuals (e.g., people, students), Groups (e.g., families, classrooms), Artifacts (e.g., books,
newspapers), Geographical units (e.g., cities, regions), Social interactions (e.g., divorces,
arrests)
Types of Research Questions
Descriptive - describe what is going on or what exists. E.g.Public opinion polls - proportion If individual scores are analyzed → the individual is the unit.If class averages are analyzed
of people who hold various opinions are primarily descriptive in nature. what percent of the → the group/class is the unit.
population would vote for a Democratic
Relational - relationships between two or more variables. poll that compares what proportion Two research fallacies - Ecological Fallacy - Drawing conclusions about individuals based
of males and females say they would vote for a Democratic only on group-level data. Eg: Assuming a student is a math whiz just because she’s from
Causal. determine whether one or more variables (e.g.a program or treatment variable) the highest-scoring math class—she might actually be the lowest scorer in that group.
affects one or more outcome variables. poll to try to determine whether a recent political Exception Fallacy - Drawing conclusions about a group based on an exceptional individual
advertising campaign changed voter preferences case.Eg: Seeing one woman make a driving mistake and concluding that all women are
bad drivers.
Time-based research designs
Hourglass Structure of Research - Start Broad: Begin with a general area of interest (e.g.,
Cross-Sectional Study - Snapshot in time: Data is collected once, at a single point. Provides a
improving math performance).Narrow Focus: Refine to a specific research question or
"slice" of the phenomenon being studied. Useful for descriptive or correlational research.
hypothesis that is testable (e.g., does a certain computer-based program improve math
Eg A survey measuring public opinion on climate change conducted in July 2025.
scores?).Data Collection and Analysis: Collect direct measurements or observations and
Longitudinal Study - multiple waves of measurement over time. Tracks changes, trends, or
analyze them.Broaden Again: Use results to generalize back to the broader topic or
developments in variables. studying causal relationships or long-term effects. Two types:
population.
Repeated Measures Design - two or a few measurement points. experimental or quasi-
experimental designs. Appropriate for repeated measures ANOVA and similar analyses.
Time Series Design - Involves many measurement points, typically 20 or more. time series Components of a Causal Study
analysis, trends, cycles, and patterns over time. economics, epidemiology, program The Research Problem - A broad issue or concern that motivates the study (e.g.,
evaluation. unemployment). Research Question - specific inquiry, framed within a theoretical context
(e.g., does supported employment help maintain jobs?). Hypothesis -A precise statement
predicting the outcome (e.g., a program will increase employment rates after 6 m). Program
Relationships in Research - how two variables are connected.
(Cause) -The intervention, event, or treatment believed to influence the outcome. Can be
Nature of the Relationship - Correlational Relationship: Two variables change in a
controlled (e.g., a program) or natural (e.g., an economic shift). Outcomes (Effect) - The
synchronized way, but one does not necessarily cause the other. Eg: Math ability and music
results or dependent var being measured (e.g., employment status, absenteeism). Units - The
proficiency may be correlated, but one doesn’t cause the other. can be misleading if
people, organizations, or entities being studied or sampled (e.g., newly employed
interpreted as causation. Causal Relationship:
individuals). Design - Refers to how units are assigned to different treatments or programs. A
One var directly causes a change in the other. harder to prove , requires experimental control.
well-constructed design ensures valid comparisons between groups (e.g., experimental vs.
E.g.Increasing the amount of daily exercise causes a reduction in blood pressure. Third
control). Concept Clarifications -Construct vs. Operationalization:A construct is the
Variable Problem - may influence both variables, giving the illusion of a direct relationship.
conceptual idea (e.g., support services). Operationalization is how that idea is measured in
E.g.: Wealthier students may use computers more and get better grades, but it’s the
reality (e.g., number of follow-up calls made). Hierarchical/Multi-level Analysis:
socioeconomic status not computer use that drives the results.
Sometimes, units exist at multiple levels (e.g., sampling cities and then families within
Patterns of Relationships - form or direction
them). Positivism vs. Post-Positivism - Epistemology: The philosophy of how we come to
No Relationship: - one var tells you nothing about the other. Eg: Lifeline length on your hand
know things. Methodology: The practice—the specific methods we use to gain knowledge.
and GPA. Positive: Higher values of one var are associated with higher values of another.
These are closely linked:
Eg: More years of education → higher income. Negative (Inverse):Higher values of one var
methodology is guided by epistemological assumptions. Positivism Dominated science in the
are associated with lower values of another. Eg: Higher self-esteem → lower paranoia in
early 20th century. Emphasizes observable, measurable phenomena. Rejects metaphysics;
psychiatric patients. Curvilinear: The direction of the relationship changes across the range reality is what we can observe. Seeks to find laws of cause and effect using the scientific
of values. Eg: A drug might reduce illness up to a point but then worsen it at high doses due method (e.g., experiments).Believes in objectivity, empiricism, and determinism. Post-
to side effects. Positivism -A rejection of the core tenets of positivism. critical realism: reality exists, but we
Variables - any characteristic or entity that can take on different values anything that can vary can’t know it perfectly.
Quantitative - (e.g., age, height, income), Qualitative/text-based (e.g., city, religion, gender). Observations are theory-laden. Emphasizes: Multiple measures and Acknowledging bias
not limited to numbers or traditional measurements. Attributes - specific value of a variable. Concept Mapping - structured group process to visually organize ideas about a
Eg: For the variable student grade, attributes could be "pass" or "fail". For agreement, topic.designed primarily for groups, by-step approach led by a facilitator. Uses multivariate
attributes might be:1 = strongly disagree…5 = strongly agree. Independent vs. Dependent statistical analyses to aggregate individual inputs into a group product. Requires specialized
Variables important in cause-effect (causal) research: Independent Variable (IV): The cause software for data processing and visualization. Helps researchers and stakeholders formulate
or what is manipulated (e.g., a new teaching method). Dependent Variable (DV): The effect research problems. strategic planning, product development, market analysis, research
or what is measured as the outcome (e.g., student test scores). Good variables have two key project design.
traits: Six Steps of Concept Mapping: Preparation Step - participants (10– 20 stakeholders).project
Exhaustive – all possible responses are covered. Eg: A religion variable should not only focus (e.g., program definition, expected outcomes). schedule Generation Step - Participants
include "Christian," "Jewish," and "Muslim" but should also have an “Other” category to generate set of statements Methods: brainstorming, focus groups, text analysis. Structuring
capture all remaining possibilities. Mutually Exclusive – a respondent should only fit into Step - participants sort statements into similar piles,rate each statement on a scale.
one attribute at a time. Bad e.g.: If someone is employed and job-hunting, they might check Representation Step -Statistical analysis of sorting and rating data: Multidimensional Scaling
both "employed" and "unemployed".In such cases, surveys often use check-all-that-apply (MDS) creates a point map positioning similar statements closer together. Cluster Analysis
questions, where each item is treated as its own dichotomous variable (e.g., “checked” = 1, groups points into clusters representing related concepts or constructs. Interpretation Step -
“unchecked” = 0). Hypothesis - specific, testable prediction about the expected outcome of a Facilitator and stakeholders collaboratively label and interpret the clusters and maps.
study, stated in concrete terms. not all research requires hypotheses— especially exploratory Utilization Step - Use the concept maps to guide research design, operationalize programs, or
or inductive studies—many do, null hypothesis (H₀): states that there is no effect or no develop measures.
relationship. alternative hypothesis (H₁ or Hₐ): states that there is an effect or relationship. Evaluation research - evaluation - Systematic assessment of the worth or merit of some
One-tailed: The prediction specifies a direction (e.g., “absenteeism will decrease”).Two- object.“Object” can mean programs, policies, technologies, people, needs, or activities.
tailed: The prediction simply anticipates a change but not its direction (e.g., “there will be a systematic and data-driven, focusing on gathering, assessing, and interpreting information
difference in depression levels”). Goals of Evaluation - provide useful feedback to sponsors, clients, staff, etc
Evaluation Strategies - Scientific-Experimental Models Prioritize objectivity, accuracy,
Types of Data - Quantitative data: rigorous and scientific, numerical in nature. Qualitative validity.Include experimental/quasi-experimental designs, cost-effectiveness, theory-driven
data: detailed and sensitive, non-numerical, including text, images, audio, video, etc. evaluation. ManagementOriented Systems Models - Focus on comprehensiveness and
Quantitative measures (like a self-esteem scale) are built on qualitative judgments (e.g., organizational context. Include PERT, CPM, Logical Framework
need to know before your exams Athabasca University
,Four types of research based on objectives.......................................... 1 Student T-test (Sample Statistical Techniques) .................................... 9
Problem Key Sections of the Paper ..................................................... 1 Choosing a Statistical Test................................................................ 10
Five Big Words in Social Research. .....................................................1 Data Mining, Knowledge Discovery (KDD), Kind of Data ....................... 10
Types of Research Questions............................................................. 1 Simulation ..................................................................................... 11
Optimization: Numerical Analysis, Numerical Methods, and Operations
Time-based research designs ............................................................. 2
Research ....................................................................................... 11 Modeling, Mathematical
Relationships in Research - how two variables are connected. ................. 2
Variables ........................................................................................ 2 Model, Bayesian ........................................... 11 Ethics, Refereeing,
Hypothesis ..................................................................................... 2 Plagiarism ......................................................... 11
Types of Data ................................................................................. 2
Surveys ....................................................................................... 12
Unit of analysis ............................................................................... 2
Correlation ................................................................................... 12
Two research fallacies ...................................................................... 2
Research Design............................................................................ 12
Hourglass Structure of Research ........................................................2
Types of Designs ........................................................................... 12
Components of a Causal Study ......................................................... 2 Research Methods in Information Sciences, Computational complexity
Positivism vs. Post-Positivism............................................................. 2 theory, Complexity and big-O notation.............................................. 13 Content Analysis,
Concept Mapping. ............................................................................2
Principal component analysis .................................. 14
Evaluation research ......................................................................... 3
Discourse Analysis, Social network, Social computing, Case Study
The Planning-Evaluation Cycle ........................................................... 3
Approach...................................................................................... 14
Study Questions .............................................................................. 3 Longitudinal Study......................................................................... 14
Types of Research Questions............................................................. 3
Action Research. ........................................................................... 15
Statistical significance ....................................................................... 3
Project Management.......................................................................15
Research Question and Hypothesis .....................................................4 Information Technology Infrastructure Library (ITIL) ........................... 15
Meta-Analysis ................................................................................. 4
Presentation, Research poster, Oral presentation, PowerPoint ................15
Sampling. ...................................................................................... 4
How to Write Research Paper, Thesis, Dissertation, References .............. 16
Probability sampling ........................................................................ 4 Qualitative Research, Measures, Debate, Data, Approaches, Methods,
Nonprobability sampling.................................................................... 4 Validity ........................................................................................ 16
Measurement – Construct Validity ......................................................4 Assignment 1 ...................................... Error! Bookmark not defined.
Types of measurement validity........................................................... 5 Assignment 2 ...................................... Error! Bookmark not defined.
The Four Levels of Measurement........................................................ 5 Four types of research based on objectives:
Descriptive – systematically describes a situation or phenomenon.
Software Quality ............................................................................. 5 Correlational – identifies relationships between variables.
Explanatory – explains why and how relationships exist.
Wechsler Adult Intelligence Scale ....................................................... 5
Exploratory – investigates little-known areas or research possibilities.
Modern Language Aptitude Test ........................................................ 5 Quantitative research (structured, preplanned, measures extent). Qualitative research
(unstructured, flexible, explores nature).
General issues in scaling .................................................................. 5
Thurstone, Likert, Gutman Scaling. ....................................................5 Problem Key Sections of the Paper:
Abstract: A brief summary of the entire study
Reliability ....................................................................................... 6 Introduction (includes Literature Review though untitled): research problem, relevant past
Validity ......................................................................................... 6 studies, rationale, research question and hypothesis
Method: Sample who participated in the study and how they were selected. It typically
True Score Theory ........................................................................... 7 includes: (sample size). Demographic details Sampling method, e.g.: Inclusion or exclusion
criteria, Measures instruments or tools used to collect data, Surveys, questionnaires, tests, or
Measurement Error .......................................................................... 7 observational checklists Descriptions of scales (e.g., Likert scales) and e.g. items. Design
Analysis ......................................................................................... 7 overall structure of the study, including: quantitative or qualitative, experimental or
observational, cross-sectional or longitudinal. independent and dependent variables, and how
Data Preparation (Analysis) .............................................................. 7 groups were assigned (randomly or not). Procedures step-by-step how the research was
conducted, such as: How participants were contacted and recruited. The sequence of
Descriptive Statistics (Analysis)......................................................... 7 activities (e.g., initial briefing, consent, data collection, debriefing).Duration Settings for data
collection, ethical considerations taken, like consent forms Results: Presents the data and
Inferential Statistics (Analysis) ..........................................................8
findings, statistical analysis, charts, or tables. Conclusions, References
T-test, Dummy Variables, GLM, Posttest-Only Analysis, Factorial Design Analysis,
Randomized Block Analysis, Analysis of Covariance, NEGD, RD
Five Big Words in Social Research
Analysis, Regression Point Analysis (Inferential statistical techniques)...... 8
Theoretical - ideas, models, or frameworks that explain how the world works. often seeks to
Event Data ..................................................................................... 9 develop, test, or refine theories.
ANOVA, Bayesian, Correlation, Factor Analysis MANOVA, Graph-Theoretic Empirical - observe and measure in the real w orld. collects data from experience, such as
Equation Modeling, Pearson Chi-squared Test, Regression Analysis, surveys, interviews, or observations. Nomoth etic - general case or universal laws rather than
,individual uniqueness. aims to find patterns that apply to groups or populations, not just one wording, cultural context). Qualitative data can be converted into quantitative form, such as
person. Contrasts with idiographic, which is concerned with the individual case. coding statements into numerical categories or similarity matrices. Eg grouping statements
Probabilistic - In the post-positivist view, research no longer claims to provide and representing similarities with a 0– 1 matrix.
certainty.Conclusions are probabilities, not absolute truths. findings are statistical estimates Unit of analysis - primary entity being studied or analyzed in a research project. what or who
with some (confidence intervals, significance levels). is being analyzed—not simply sampled—depends on how the data is analyzed, not just
Causal - cause-and-effect relationships—ultimate goal in much research is to determine how collected.
changes (interventions, programs, treatments) lead to changes in outcomes. Individuals (e.g., people, students), Groups (e.g., families, classrooms), Artifacts (e.g., books,
newspapers), Geographical units (e.g., cities, regions), Social interactions (e.g., divorces,
arrests)
Types of Research Questions
Descriptive - describe what is going on or what exists. E.g.Public opinion polls - proportion If individual scores are analyzed → the individual is the unit.If class averages are analyzed
of people who hold various opinions are primarily descriptive in nature. what percent of the → the group/class is the unit.
population would vote for a Democratic
Relational - relationships between two or more variables. poll that compares what proportion Two research fallacies - Ecological Fallacy - Drawing conclusions about individuals based
of males and females say they would vote for a Democratic only on group-level data. Eg: Assuming a student is a math whiz just because she’s from
Causal. determine whether one or more variables (e.g.a program or treatment variable) the highest-scoring math class—she might actually be the lowest scorer in that group.
affects one or more outcome variables. poll to try to determine whether a recent political Exception Fallacy - Drawing conclusions about a group based on an exceptional individual
advertising campaign changed voter preferences case.Eg: Seeing one woman make a driving mistake and concluding that all women are
bad drivers.
Time-based research designs
Hourglass Structure of Research - Start Broad: Begin with a general area of interest (e.g.,
Cross-Sectional Study - Snapshot in time: Data is collected once, at a single point. Provides a
improving math performance).Narrow Focus: Refine to a specific research question or
"slice" of the phenomenon being studied. Useful for descriptive or correlational research.
hypothesis that is testable (e.g., does a certain computer-based program improve math
Eg A survey measuring public opinion on climate change conducted in July 2025.
scores?).Data Collection and Analysis: Collect direct measurements or observations and
Longitudinal Study - multiple waves of measurement over time. Tracks changes, trends, or
analyze them.Broaden Again: Use results to generalize back to the broader topic or
developments in variables. studying causal relationships or long-term effects. Two types:
population.
Repeated Measures Design - two or a few measurement points. experimental or quasi-
experimental designs. Appropriate for repeated measures ANOVA and similar analyses.
Time Series Design - Involves many measurement points, typically 20 or more. time series Components of a Causal Study
analysis, trends, cycles, and patterns over time. economics, epidemiology, program The Research Problem - A broad issue or concern that motivates the study (e.g.,
evaluation. unemployment). Research Question - specific inquiry, framed within a theoretical context
(e.g., does supported employment help maintain jobs?). Hypothesis -A precise statement
predicting the outcome (e.g., a program will increase employment rates after 6 m). Program
Relationships in Research - how two variables are connected.
(Cause) -The intervention, event, or treatment believed to influence the outcome. Can be
Nature of the Relationship - Correlational Relationship: Two variables change in a
controlled (e.g., a program) or natural (e.g., an economic shift). Outcomes (Effect) - The
synchronized way, but one does not necessarily cause the other. Eg: Math ability and music
results or dependent var being measured (e.g., employment status, absenteeism). Units - The
proficiency may be correlated, but one doesn’t cause the other. can be misleading if
people, organizations, or entities being studied or sampled (e.g., newly employed
interpreted as causation. Causal Relationship:
individuals). Design - Refers to how units are assigned to different treatments or programs. A
One var directly causes a change in the other. harder to prove , requires experimental control.
well-constructed design ensures valid comparisons between groups (e.g., experimental vs.
E.g.Increasing the amount of daily exercise causes a reduction in blood pressure. Third
control). Concept Clarifications -Construct vs. Operationalization:A construct is the
Variable Problem - may influence both variables, giving the illusion of a direct relationship.
conceptual idea (e.g., support services). Operationalization is how that idea is measured in
E.g.: Wealthier students may use computers more and get better grades, but it’s the
reality (e.g., number of follow-up calls made). Hierarchical/Multi-level Analysis:
socioeconomic status not computer use that drives the results.
Sometimes, units exist at multiple levels (e.g., sampling cities and then families within
Patterns of Relationships - form or direction
them). Positivism vs. Post-Positivism - Epistemology: The philosophy of how we come to
No Relationship: - one var tells you nothing about the other. Eg: Lifeline length on your hand
know things. Methodology: The practice—the specific methods we use to gain knowledge.
and GPA. Positive: Higher values of one var are associated with higher values of another.
These are closely linked:
Eg: More years of education → higher income. Negative (Inverse):Higher values of one var
methodology is guided by epistemological assumptions. Positivism Dominated science in the
are associated with lower values of another. Eg: Higher self-esteem → lower paranoia in
early 20th century. Emphasizes observable, measurable phenomena. Rejects metaphysics;
psychiatric patients. Curvilinear: The direction of the relationship changes across the range reality is what we can observe. Seeks to find laws of cause and effect using the scientific
of values. Eg: A drug might reduce illness up to a point but then worsen it at high doses due method (e.g., experiments).Believes in objectivity, empiricism, and determinism. Post-
to side effects. Positivism -A rejection of the core tenets of positivism. critical realism: reality exists, but we
Variables - any characteristic or entity that can take on different values anything that can vary can’t know it perfectly.
Quantitative - (e.g., age, height, income), Qualitative/text-based (e.g., city, religion, gender). Observations are theory-laden. Emphasizes: Multiple measures and Acknowledging bias
not limited to numbers or traditional measurements. Attributes - specific value of a variable. Concept Mapping - structured group process to visually organize ideas about a
Eg: For the variable student grade, attributes could be "pass" or "fail". For agreement, topic.designed primarily for groups, by-step approach led by a facilitator. Uses multivariate
attributes might be:1 = strongly disagree…5 = strongly agree. Independent vs. Dependent statistical analyses to aggregate individual inputs into a group product. Requires specialized
Variables important in cause-effect (causal) research: Independent Variable (IV): The cause software for data processing and visualization. Helps researchers and stakeholders formulate
or what is manipulated (e.g., a new teaching method). Dependent Variable (DV): The effect research problems. strategic planning, product development, market analysis, research
or what is measured as the outcome (e.g., student test scores). Good variables have two key project design.
traits: Six Steps of Concept Mapping: Preparation Step - participants (10– 20 stakeholders).project
Exhaustive – all possible responses are covered. Eg: A religion variable should not only focus (e.g., program definition, expected outcomes). schedule Generation Step - Participants
include "Christian," "Jewish," and "Muslim" but should also have an “Other” category to generate set of statements Methods: brainstorming, focus groups, text analysis. Structuring
capture all remaining possibilities. Mutually Exclusive – a respondent should only fit into Step - participants sort statements into similar piles,rate each statement on a scale.
one attribute at a time. Bad e.g.: If someone is employed and job-hunting, they might check Representation Step -Statistical analysis of sorting and rating data: Multidimensional Scaling
both "employed" and "unemployed".In such cases, surveys often use check-all-that-apply (MDS) creates a point map positioning similar statements closer together. Cluster Analysis
questions, where each item is treated as its own dichotomous variable (e.g., “checked” = 1, groups points into clusters representing related concepts or constructs. Interpretation Step -
“unchecked” = 0). Hypothesis - specific, testable prediction about the expected outcome of a Facilitator and stakeholders collaboratively label and interpret the clusters and maps.
study, stated in concrete terms. not all research requires hypotheses— especially exploratory Utilization Step - Use the concept maps to guide research design, operationalize programs, or
or inductive studies—many do, null hypothesis (H₀): states that there is no effect or no develop measures.
relationship. alternative hypothesis (H₁ or Hₐ): states that there is an effect or relationship. Evaluation research - evaluation - Systematic assessment of the worth or merit of some
One-tailed: The prediction specifies a direction (e.g., “absenteeism will decrease”).Two- object.“Object” can mean programs, policies, technologies, people, needs, or activities.
tailed: The prediction simply anticipates a change but not its direction (e.g., “there will be a systematic and data-driven, focusing on gathering, assessing, and interpreting information
difference in depression levels”). Goals of Evaluation - provide useful feedback to sponsors, clients, staff, etc
Evaluation Strategies - Scientific-Experimental Models Prioritize objectivity, accuracy,
Types of Data - Quantitative data: rigorous and scientific, numerical in nature. Qualitative validity.Include experimental/quasi-experimental designs, cost-effectiveness, theory-driven
data: detailed and sensitive, non-numerical, including text, images, audio, video, etc. evaluation. ManagementOriented Systems Models - Focus on comprehensiveness and
Quantitative measures (like a self-esteem scale) are built on qualitative judgments (e.g., organizational context. Include PERT, CPM, Logical Framework
