Research Ethics and Experimental Design

Ethics in Research

• Ethics are guidelines designed to aid researchers in conducting ethical research.
• Government funding of scientific studies can be influenced by congressional priorities, shaping the focus (e.g., AIDS research).
• Corporate funding, particularly in areas like drug research, may lead to conflicts of interest and erode public trust in science.

Variables in Research

• Independent variables are manipulated by the researcher and presumed to cause changes in another variable; for example, comparing therapy to no therapy.
• Dependent variables are the presumed effect or outcome, measured by the researcher, and influenced by the independent variable.
• Confounding variables, also known as extraneous variables, compete with the independent variable in explaining the dependent variable.

Experimental Designs

• Within participant designs expose all participants to all treatments at different times or in different sequences (repeated measures design).
• Between participant designs use random assignment to expose different groups to different levels of the independent variable (randomized designs).
• Posttest-only control group design represents the most basic between participant design, featuring one independent and one dependent variable without a pretest.
• Mixed designs combine within and between participant approaches, using at least one of each type of independent variable (e.g., pretest-posttest control group design).
• Factorial designs study two or more independent variables to determine their separate and combined effects on the dependent variable.

Types of Research

• Experimental research seeks to establish cause-and-effect relationships through controlled psychological experiments.
• Nonexperimental research focuses on describing phenomena, events, or situations without manipulating variables.
• Quantitative studies collect numerical data, which provides a numerical analysis of the research questions.
• Qualitative studies collect non-numerical data to answer research questions, to achieve a deeper understanding of the research questions.

Qualitative Research

• Strengths include conducting studies in natural settings, describing common identity, and advancing theoretical understanding of phenomena.
• Weaknesses include difficulty generalizing findings, potential researcher disagreement, and the absence of objective hypothesis testing.

Variables and Their Measurement

• A variable is a condition or characteristic that can take on different values or categories, (e.g., gender.)
• Measurement of a variable is assigning symbols or numbers according to a set of rules, for example, male/female for gender, reaction time in minutes or seconds.
• Quantitative variables vary by degree or amount, such as reaction time, height, or anxiety levels.
• Categorical variables vary by type or kind, such as gender, religion, or method of therapy.

Experimental Research Settings

• Field experiments are conducted in real-life settings.
• Laboratory settings are conducted in controlled laboratory environments.
• Internet experiments are conducted over the internet.
• A strength of experiments lies in the manipulation of the independent variable.
• Correlational/non-experimental studies cannot establish causal relationships.
• Cross-sectional designs collect data at a single point in time.
• Longitudinal designs collect data at multiple points in time.

Hypothesis Testing

• Hypothesis testing involves evaluating a predicted relationship or hypothesis by comparing observations with the hypothesis.
• A statistical hypothesis is a precise, quantitative statement translating the research hypothesis into a testable form using statistical methods.
• A research hypothesis is a theoretical statement outlining the expected relationship between variables based on existing understanding.

Methods of Data Collection

• Common data collection methods include tests, questionnaires, and interviews.

Tests

• Used to measure personality, aptitude, achievement, and performance, with information on reliability, validity, and norms.
• Strengths of tests include standardization, strong psychometric properties, group administration, and ease of quantitative data analysis.
• Weaknesses include participant reactivity, potential bias, and lack of open-ended questions.

Questionnaires

• A self-report instrument completed by research participants.
• Strengths include providing insights into participants' perspectives and allowing both close-ended and open-ended items.
• Weaknesses include participant reactivity, low response rates, and potential lack of self-awareness.

Interviews

• A structured conversation where an interviewer asks an interviewee a series of questions.
• Strengths include suitability for measuring attitudes, probing capabilities, and provision of subjective perspectives.
• Weaknesses include high cost and time consumption, potential investigator effects, and low perceived anonymity.

Focus Groups

• A moderator guides a small, homogeneous group in discussing a research topic or issue.
• Focus groups typically consist of 6-12 people.
• Sessions generally last 1-3 hours.
• Recording methods include audio and/or video recordings.
• They are not group interviews.

Observation

• Naturalistic observation takes place in real-world settings.
• Laboratory observation is conducted in a lab or other controlled environment.
• Time-interval sampling involves observation during pre-selected time intervals.
• Event sampling involves observing every instance a particular event occurs.

Strengths of Observation

• Provides firsthand experience.
• Offers relatively objective behavior measurement with moderate realism.
• Observers may notice details unnoticed by those in the setting.
• Helps understand the importance of contextual factors.
• Suitable for participants with weak verbal skills.

Weaknesses of Observation

• Reasons for observed behavior can be unclear.
• Reactivity may occur when respondents know they are being observed.
• Investigator effects may influence observations.
• Observation is unsuitable for large or dispersed populations.
• Certain settings and content of interest are unobservable.

Existing or Secondary Data

• Data initially used for a purpose other than the current research study.
• Types of data include documents, physical data, and archived research data.
• Archived research data is secondary data collected by other researchers for a different purpose.

Strengths of Documents and Physical Data

• Unobtrusive nature.
• Can be collected for past time periods.
• Provides background/historical context.
• Useful for corroboration.
• Grounded in local settings.
• Useful for exploration.
• Can be unobtrusive

Weaknesses of Documents and Physical Data

• May be incomplete.
• Might represent only one viewpoint.
• Access could be restricted.
• May not provide insight into participants' personal thoughts (for physical data).
• May not apply to general populations.

Weaknesses of Archived Research Data

• May not be available for the population of interest.
• May not be available for specific research questions.
• Data may be outdated.
• Open-ended or qualitative data is usually unavailable.
• Most important findings may already be extracted.

Evaluating Web Pages

• A validity of information is a factor.
• Authority check author and their credentials.
• Accuracy check credentials and email address for contact.
• Objectivity identify any evidence of bias.
• Currency ensure regular updates to web page.
• Coverage ensure information can be viewed without fees or extra software.

Research Questions

• A research question identifies the variables and relationships to be studied.
• Good research questions should express a relationship between variables and they should be empirically testable.
• Specificity of research questions ensures researcher clarity.
• Formulating hypotheses involves making the best prediction or a tentative solution to a problem based on a literature review.

Ethical Considerations

• An Institutional Review Board (IRB) evaluates the ethical aspects of research.
• Informed consent includes details such as investigator information, research purpose, expectations, voluntary participation statement, privacy, IRB approval, contact for questions/complaints, date, and signature.
• Active deception involves giving false information.
• Passive deception includes withholding information.

Scales of Measurement

• Nominal scales classify or categorize data using symbols, for example gender, ethnicity or religion.
• Ordinal scales rank-order data but intervals may not be equal, for example finishing order in a race or letter grades.
• Interval scales include equal distances between adjacent numbers, for example temperature in Celsius.
• Ratio scales have all properties of other scales plus an absolute zero point, for example weight or height.

Reliability and Validity

• Reliability refers to the consistency or stability of measurement scores.
• Validity refers to the extent to which a procedure measures what it is supposed to measure, with correct score interpretation.