Foundations of Research Methods — Comprehensive Study Notes
What is Scientific Research?
- Scientific research is the systematic investigation of scientific theories and hypotheses.
- It is a process of rigorous reasoning based on interactions among theories, methods, and findings.
- It builds on understanding derived from objective testing of models or theories.
- The accumulation of scientific knowledge is laborious, plodding, circuitous, and indirect.
- Scientific knowledge is developed and improved through critique, contested findings, replication, and convergence.
- Scientific knowledge is developed through sustained efforts.
- Scientific inquiry must be guided by fundamental principles.
Fundamental Principles
I. Ask significant questions that can be answered empirically
- The formulation of a problem is often more essential than its solution, which may be merely a matter of mathematical or experimental skill. To raise new questions, new possibilities, to regard old questions from a new angle, requires creative imagination and marks real advance in science (Einstein & Infeld, 1938).
- The research questions must be asked in a way that allows empirical investigation.
II. Link research to relevant theory
- Scientific research can be guided by a conceptual framework, model, or theory that generates questions to be asked or answers to the questions posed.
- Theory drives the research question, the use of methods, and the interpretation of results.
III. Select and apply research designs and methods that permit direct investigation of the question
- Trustworthiness of a study is predicated on major elements:
- Suitability of the proposed design or methodology to address the specific questions posed.
- Scientific rigor by which the methodology is applied.
- The link between question and methodology must be clear and justified.
- Detailed description of the method, measures, data collection procedures, data analyses, and subjects must be available to permit replication.
IV. Provide a coherent and explicit chain of reasoning that can be replicated
- What assumptions underlie the inferences, and were they clearly stated and justified?
- How was evidence judged to be relevant?
- How were alternative, competing hypotheses identified, considered, and accounted for (accepted or discarded)?
V. Replicate and generalize across studies
- Internal Validity: Observations are consistent and generalizable across observers, tasks, and measurement occasions.
- Statistical methods e.g., correlation;
- Non-statistical methods e.g., triangulation, comparative analysis.
- External Validity: Extent to which treatment conditions and participant population reflect the world to which generalization is desired.
VI. Report research publicly to encourage professional scrutiny, critique and replication
- Criticism is essential to scientific progress.
- The extent to which new findings can be reviewed, contested, and accepted or rejected depends on accurate, comprehensive, and accessible records of:
- Data
- Methods
- Inferential reasoning
Common Conceptions/Misconceptions About Scientific Research
- ✓ Experimental research is more "scientific" than descriptive or qualitative research.
- A study is deemed to be "scientific" when:
- There are clear, testable questions underlying the design.
- The methods are appropriate to answer the questions and falsify competing hypotheses.
- The study is explicitly linked to theory and previous research.
- The data are analyzed systematically and with appropriate tools.
- The data are made available for review and criticism.
- ✓ Research in education is fundamentally different from the hard sciences.
Basic Elements of the Scientific Methods
- Scientific method refers to a standardized set of techniques for building scientific knowledge, such as:
- how to make valid observations,
- how to interpret results,
- and how to generalize those results.
- 1. Replicability: Replicate or repeat a scientific study and obtain similar, if not identical, results.
- 2. Precision: Theoretical concepts defined so that others can measure those concepts and test the theory.
- 3. Falsifiability: Theories that cannot be tested or falsified are not scientific theories; such knowledge is not scientific knowledge.
- 4. Parsimony: When there are multiple explanations, accept the simplest or most economical explanation.
Types of Scientific Research
- Scientific Research categories: Exploratory, Descriptive, Explanatory.
- 1. Quantitative Research: data are numeric, often associated with the positivist/post-positivist paradigm; aims to enable statistical calculations and generalizable conclusions.
- 2. Qualitative Research: associated with the social constructivist paradigm; emphasizes the socially constructed nature of reality, aims to uncover deeper meanings and significance of human behavior and experience; seeks rich, comprehensive understanding rather than broad generalization.
Scientific Method of Research: Four Steps
- 1. Observation
- 2. Hypothesis
- 3. Testing
- 4. Predictions
- The cycle involves Induction (from Observation to Theory) and Deduction (theory guiding observations).
- If the experiments prove the hypothesis true, it becomes a theory or law of nature.
- If the experiments prove the hypothesis false, the hypothesis must be rejected or modified.
- Proper use of the method should yield predictive power for phenomena not yet tested.
Observation
- An act of recognizing and noting a fact or occurrence, often involving measurement with instruments.
- Experiments performed in the laboratory; experiments drawn from the literature.
- Keys to implementation in research:
- Sorting Observations (from Literature Searches):
- Useful
- Contains unanswered questions you think you can address
- Not Useful (Yet!)
- Formulate Hypothesis.
- Note results (in report format—optional) for future use in confirming/denying your hypothesis once it is found.
Hypothesis
- Hypothesis: a tentative assumption made to draw out and test its logical or empirical consequences.
- Good Hypothesis can be TESTED with Experiment or Calculation.
- Requires substantial thought and reading—leap from observation to hypothesis; worth the effort because it prepares you for laboratory work.
- Suggestions for developing hypotheses:
- Couple archival journal reading with theory.
- Couple with common sense or intuition about how things should be in the system studied.
- Follow logical reasoning about what you read with scientific and/or mathematical basis (use drawings/diagrams).
- Write up your thoughts and opinions in report format or in a notebook.
- Hypothesis shapes the study in terms of:
- Identification of study sample size
- Issues in data collection
- Proper analysis of the data
- Data interpretation
- Traditional framing:
- Ho: "Null" hypothesis (assumed)
- H1: "Alternative" hypothesis
- Examples:
- Ho: There is no association between the exposure and disease of interest
- H1: There is an association between the exposure and disease of interest (beyond what might be expected from random error alone)
Experimental Testing
- To apply a test as a means of analysis or diagnosis.
- Keys to implementation:
- Good TESTS will prove or disprove your hypothesis.
- Experimental tests can be performed within computing (e.g., Coventor model with calculations and predictions).
- Consider all alternatives; an experiment may not disprove all aspects of the hypothesis, but may disprove parts; note which aspects the experiment tests.
- Consider the availability of instrumentation to perform your tests.
Predictions
- Definition: To declare or indicate in advance; especially foretell on the basis of observation, experience, or scientific reason.
- Keys to implementation:
- Good Predictions can be tested against your hypothesis.
- Revisit basics (textbook theory) to develop a mathematical model/construct to help make predictions about more systems than you can reasonably test.
- Suggestions:
- Begin thinking about predictions after you have a hypothesis; however, if the hypothesis is proven false, the prediction will also fail.
- Use a mathematical model to test your prediction.
Mathematical Models
- Empirical vs. Deterministic
- Empirical: Based on experimental observation.
- Deterministic: Based on first principles (e.g., Density Functional Theory, Molecular Dynamics).
- Keys to implementation:
- All thesis experiments should have empirical models (at least).
- If you do not have a mathematical model—only data—then you need to find an equation for least-squares fitting of your data.
- If you encounter an interesting hypothesis that can be tested via deterministic models (first principles), consult with modeling to discuss the hypothesis and modeling.
Understanding of Methodology and Methods
- Methodology is defined as:
- "the analysis of the principles of methods, rules, and postulates employed by a discipline";
- "the systematic study of methods that are, can be, or have been applied within a discipline"; or
- "a particular procedure or set of procedures".
- It can be viewed as: a collection of theories, concepts or ideas; comparative study of different approaches; critique of the individual methods.
- Key point: Methodology refers to more than a simple set of methods; it refers to the rationale and philosophical assumptions that underlie a particular study.
Method
- Method is a (systematic?) codified series of steps taken to complete a task or reach a specific objective.
- The scientific method attempts to minimize the influence of the researchers' bias on the outcome of an experiment.
- The researcher may have a preference for one outcome or another; it is important that this preference not bias the results or their interpretation.
- Cautions:
- Common sense and logic can tempt us to think no test is needed.
- It is a mistake to ignore or rule out data that do not support the hypothesis.
- There is no single universal formal method; there are several variants, and researchers tailor the process to the problem and their working methods.
- Methodological choice: Scientist vs Engineer
- A scientist asks "why?" and researches the answer to the question.
- An engineer asks "how" to solve it and "how" to implement that solution, or how to improve it if a solution exists.
- Core idea: scientists build in order to learn; engineers learn in order to build.
Quantitative vs Qualitative Methods
- Quantitative methods and scientific rigor:
- Data are numeric, collected with standardized methods, replicable, and analyzable with sophisticated statistical techniques.
- Qualitative methods and scientific rigor:
- Many qualitative approaches contend that there is no objective social reality; knowledge is constructed by observers influenced by traditions, beliefs, and social/political environments.
Four issues affecting the choice of method
- 1. Credibility of findings
- 2. Staff skills
- 3. Costs
- 4. Time constraints
Applications of Methods
- Quantitative Methods
- Qualitative Methods
- Mixed Methods
Sample Size and Statistics
- Defining the sample size from population:
- Where:
- n = the desired sample size (when population is greater than 10,000)
- t = the standard normal deviate set at 2.58, which corresponds to the 99 percent confidence level (P < 0.01)
- p = the proportion in the target population estimated to have a particular characteristic (set at 0.5)
- q = 1.0 - p
- d = Degree of accuracy desired, set at 0.05
- Calculation example:
- n = rac{(2.58)^2 imes (0.50) imes (0.50)}{(0.05)^2} = 665
Class Activity
- Group activity: Using the four steps in the scientific method, design an outline to show how the scientific method can be applied to answer scientific questions.
- Email group responses to the instructor: johnk@stjohns.edu