Please enable JavaScript.
Coggle requires JavaScript to display documents.
Quantitative Research Lecture 1 - Coggle Diagram
Quantitative Research Lecture 1
Variables
Role of Variables
Predictor/independent variable = Variable used to estimate, forecast or predict future events or outcomes. Variable that is believed to influence or cause changes in another variable (the outcome)
Outcome - Variable of primary interest, Research question/hypothesis describes, explains, or predicts changes in it
Mediator - Explains or provides a link between predictor and
outcome
Confounder - A confounding variable is an external factor that is related to both the predictor and outcome variables and can potentially distort the observed relationship between them. If not controlled for, confounders can lead to incorrect conclusions about the causal relationship
Collider - A collider variable is a variable that is influenced by two or more other variables. It occurs when two or more variables cause a third variable. Importantly, conditioning on a collider can introduce bias, known as collider bias or selection bias.
Examples
Predictor - In a study examining the effect of exercise on weight loss, the amount of exercise is the predictor variable.
Outcome Variable - In the same study on exercise and weight loss, the amount of weight lost is the outcome variable.
Mediator Variable - If a study finds that exercise leads to weight loss by increasing metabolism, then increased metabolism is a mediator variable between exercise (predictor) and weight loss (outcome).
Confounding variable - In the study on exercise and weight loss, diet could be a confounding variable. If people who exercise also tend to eat healthier, then the observed weight loss might be partly due to diet rather than just exercise.
Collider - Suppose there's a study on the relationship between exercise and cardiovascular health, where physical fitness is considered. Physical fitness could be a collider because it is influenced by both exercise and genetic predisposition. If we only consider people with high physical fitness, we might distort the true relationship between exercise and cardiovascular health, as we are implicitly conditioning on the collider.
Quant vs Qual
Quantitative
numbers & statistics are being used
(answers the WHAT?)
Uses a narrative description from open-ended
interview
Description of a situation take on dynamic and (researcher’s) personal account of what is occurring – certain level of bias
Unique story of individual or group – lacks generalizability as story might not be applicable to other people
Strength is that it focuses on the why
Qualitative
descriptive styles (words/text) used
(answers the WHY?)
Tests hypothesis with closed ended surveys. Use of statistical tests. Confirms relationship and removes bias.
Must define what is being measured (limit definitions)
Deductive – broad hypothesis to specific conclusions
Strengths
Use of objective based measures - limits bias. No personal bias or perception interfering with results
Generalizable if done correctly (big sample, random sample selection, operationalized variables studied)
Deductive reasoning - general hypothesis to specific conclusions
Random Sampling - representative of general population
Quantify problems - numbers and statistics
Numerical projectiosn of perceptions - rating scales
Objective -based research - accurate conclusions that are generalizable
Advanced statistical analyses - Tests existence of statistical relationships.
Graphic data representation - figures charts table etc.
Mixed Methods
Aspects of High Quality Research
Determinism - events have natural causes (vs supernatural)
Empiricism - reliance on real evidence to confirm or refute claims.
Replicability - findings are reliable and reproducible. re-conducting prev work.
Falsifiability - hypothesis can be proved wrong
Parsimony
Experimental vs Non-Experimental
Experimetal Reserach
Researcher able to manipulate predictor variable and
subjects to identify cause-and-effect relationship
Non-experimental Research
Many scientific questions cannot be
addressed with experimental research
Ethical reasons
Practicality
Researcher cannot control, manipulate or alter predictor variable or subjects à rely on natural observations to come to conclusions
True cause-and effect relationship cannot be demonstrated
Cannot be experimental if predictor variable cannot be manipulated - eg. How violent are men and women. Gender cannot be switched. Rather obs study.
What is Research?
Careful and detailed study into a specific problem, concern, or issue using the scientific method
Scientific Method
Logically stepped process from beginning to end.
Step 3: Form a Hypothesis - prediction. IF/THEN type of statement.
Step 4: Collect data and test hypothesis.
Step 2: Background research - Helps inform what has been done before, how it was done, what gaps exist and what still needs to be done
Step 5: Analyse data - try make sense of data. Is hypothesis correct? Yes = hypothesis no rejected. No = hypothesis rejected.
Step 1: Ask a question - problem/issue that exist. Question on issue.
Step 6: Reporting -
Key Terms
Concept - Abstract thinking or idea e.g. substance abuse
Construct - Theoretical definition of a concept, must be observable or measurable e.g. alcoholism
Variable - Measured characteristics that vary/change e.g. alcohol use
Operationalization - Specifying how the variable is observed or measured e.g. AUDIT scale
Operationalizing variables
Operationalizing variables refers to the process of defining how abstract concepts, known as variables, will be measured and observed in a specific study. It involves turning theoretical constructs into measurable observations. This step is crucial for empirical research because it ensures that the variables being studied are defined in a clear and consistent manner, making it possible to collect and analyze data.
Conceptual Question
Does exercise reduce depression?
Operational definition
How will exercise be assessed? what type of exercise are you referring to? what intensity and how long per session
Depression - how will it be assessed - scale?
psychiatric exam
Specifies the way in which variable is observed
or measured