Empirical, Systematic, & Replicable Research

• Research must be systematic, empirical, and replicable to be valid.

Systematic Research

• Uses structured approaches like surveys, experiments, or content analysis for data collection.

Empirical Research

• Grounded in evidence, relying on observations, data, or existing literature to support claims.

Replicable Research

• Uses transparent and repeatable methods, allowing validation by others, and ensuring reliability.
• Scientific research differs from other ways of knowing by its heavy reliance on empirical evidence, systematic observation, and testable explanations.
• A strong research question is clear, focused, specific, relevant, researchable, and complex.
• Research questions are phrased as questions, while hypotheses are declarative statements tested through research.
• Communication research questions examples are "How does social media impact mental health?" and "How do communication strategies influence environmental advocacy?".
• Observation acts as a direct window into reality, allowing researchers to gather firsthand data about behaviors and interactions as they naturally occur.
• Researchers must consider generalizability to ensure findings are relevant and useful in real-world situations, crucial for making inferences about larger populations from smaller samples.
• Qualitative data is descriptive and non-numerical, while quantitative data is measurable and expressed in numbers for statistical analysis.
• Ethical considerations are important in research.

Theory and Observation

• Theories provide structure and direction for research by predicting and explaining phenomena.
• Empirical data, like experiments and field studies, test the theory.
• Observations refine/evolve theories, which then guide new observations and investigations.

Reasoning

• Induction develops a general theory from specific observations.
• Deduction tests a theory or hypothesis with specific observations.
• Abduction explains surprising observations by formulating new hypotheses or theories.
• Research questions are open-ended and exploratory while hypotheses are predictive and testable statements about variable relationships.
• Operationalizing constructs defines abstract concepts for measurement and observation. Example: defining "audience engagement" as likes, shares, and comments.

Research Purposes

• Exploration investigates uncharted areas.
• Description documents phenomena in detail.
• Explanation understands causal relationships.
• Prediction forecasts future behaviors.
• Literature reviews analyze existing research to refine research questions and build upon established knowledge.

Impact of Worldview

• Worldviews shape the research process.
• Worldviews influence research questions, provide methodological guidance (quantitative, qualitative, or mixed methods), and affect the interpretation of results and ethical practices.
• They also allow researchers to adapt their approach based on the research question.

Ethics in Research

• Researchers have ethical responsibilities regarding participant treatment, research design, and reporting results.

The Belmont Report

• It emphasizes respect for persons, beneficence, and justice.
• Autonomy: Individuals should be treated with respect; those with diminished autonomy are entitled to protection.
• Beneficence: Efforts should be made to secure human subject well-being, maximize possible benefits, and minimize potential harm.
• Justice: Benefits and risks of research must be distributed fairly.

Institutional Review Boards (IRBs)

• They are responsible for reviewing and approving research to ensure ethical compliance, assessing risks and benefits.
• The IRB process includes application submission, peer review, and continuous monitoring.
• The Stanford Prison Experiment lacked informed consent, had long-term effects, and lacked intervention when harm occurred.
• The Tuskegee Syphilis Study lacked informed consent, withheld treatment, exploited a vulnerable population, and failed to minimize harm.

Key Ethical Issues

• Informed consent, confidentiality, anonymity, debriefing after deceptive studies, and ensuring participant well-being.
• Informed consent requires full transparency and subjects' choice, including information, comprehension, and voluntariness.

Historical Ethical Violations

• They emphasize the importance of ethical standards. The Nuremberg Code: Developed after the Nuremberg trials, focuses on the rights and welfare of human research participants.
• Declaration of Helsinki: A set of ethical principles for medical research involving human subjects.
• Common Rule: Federal regulations for ethical conduct of human subject research in the United States.
• APA Ethical Principles: Guidelines for the professional conduct of psychologists.

Literature Reviews

• Literature reviews provide background and context, identify gaps, avoid duplication, and guide design and methodology.
• Primary research provides original, firsthand data, while secondary research offers analyses, interpretations, or summaries.
• Scholarly publications are peer-reviewed with detailed methods and results whereas popular publications are aimed at general audiences and may be less reliable.
• Boolean search strategies use AND (narrow), OR (expand), and NOT (exclude) to refine academic research.

Theories Guiding Research Questions

• Uses and gratifications Theory: People use media for emotional regulation, affecting stress.
• Social Identity Theory: We understand ourselves based on who we surround ourselves with.
• Health Belief Model (HBM): People act to prevent illness if they perceive susceptibility, seriousness, and benefits outweighing barriers.

Measurement Levels

• Measurement levels impact statistical analysis. Nominal: Categories without order. Ordinal: Ranked with unequal intervals. Interval: Equal intervals with no true zero. Ratio: Equal intervals with a true zero.

Validity and Reliability

• Reliability is consistency (same results every time), while validity is accuracy (measures what it claims to measure).
• Content validity checks if the test content represents the intended measurement area.
• Construct validity assesses if the test measures the intended theoretical construct.
• Criterion validity examines how well test scores correlate with an external measure.
• Test-retest reliability involves administering the same test at different times.
• Intercoder reliability: Two or more independent coders should assign similar ratings when evaluating the same data.

Scales

• Likert scale: Measures agreement levels.
• Semantic Differential Scale: Measures intensity of concepts using opposing adjectives.

Sampling Methods

• Probability samples are statistically representative while non-probability samples are based on researcher judgment.
• Simple random sampling randomly selects participants, stratified sampling divides the population into homogeneous groups, and cluster sampling selects entire clusters.
• Systematic sampling selects members at regular intervals.
• Convenience sampling is based on ease of access.
• Quota sampling ensures proportional representation.
• Snowball sampling: Participants recruit others.
• Purposive sampling: A researcher selects specific participants.

Standard Deviation

• The standard deviation quantifies the amount of variation or dispersion of a set of data values from the mean or average value of the dataset.
• The measures of variability are range variance standard deviation and IQR
• Higher confidence levels require larger sample sizes.

Common confidence levels

• Common confidence levels: 90% (z = 1.45), 95% (z = 1.9), 99% (z = 2.57)
• Selection bias occurs when the sample is not representative.
• Self-selection bias occurs when only certain types of people choose to participate.
• Nonresponse bias occurs when people who don't respond differ from those who do.
• Convenience sampling bias: The sample is easy to reach rather than representative.
• Sampling frame bias: The list used to select participants is incomplete or biased.

Survey Design

• Surveys can gather large numbers of responses rapidly, especially for formatted questions and with internet-based surveys.

Different Survey Designs

• Cross sectional surveys capture data at one point in time.
• Longitudinal studies track people’s changes over time.
• Panel studies: a group of individuals is sampled and recruited, and the same individuals are retained to answer questions over time.
• Trend studies: measure the same items over time using the same questions but drawing different samples from the population each time.
• Cohort studies: groups of people defined, most typically, by having an event in common.

Bias

• Social desirability bias occurs when survey respondents answer according to expectations, not beliefs.
• Question wording and sequence influence responses.
• Leading questions lead to a particular answer.
• Double-barreled questions ask two questions simultaneously.
• Negatively worded questions can confuse participants.

Increasing Response Rates

• Strategies for increasing response rates in surveys keeping surveys concise and relevant, offering incentives, clearly explaining the survey purpose, personalizing invitations, sending timely reminder emails.

Question Order

• Use the inverted funnel method, specific to broad.

Explore the key characteristics of valid research: systematic methods, empirical evidence, and replicable processes. Learn how scientific research differs from other forms of inquiry, emphasizing empirical data and testable explanations. Understand the importance of clear, focused research questions and the distinction between questions and testable hypotheses.