Scientific Study of Behavior

Scientific Study of Behavior

• Acquiring knowledge can be done through intuition, authority, or the scientific method.
• Intuition involves relying on personal experiences and opinions, often feeling emotionally right.
• Authority involves trusting experts due to their perceived knowledge.
• Using both intuition and authority lacks critical thinking and can be susceptible to bias, potentially leading to incorrect or harmful information.
• The scientific method avoids authority or intuition without evidence, employing skepticism by questioning information and seeking evidence.
• Structured and systematic observations form the basis of empiricism, a scientific approach to gaining knowledge.
• The scientific method provides an objective and systematic way to collect, evaluate, and report data, reducing biases.
• Empirical questions addressable through structured observation are posed and should be falsifiable for assertions of reality.
• Psychological research aims to describe behavior by identifying characteristics, predict behavior by determining when it will occur, and determine the causes of behavior by covariation, temporal precedence, and eliminating alternatives.
• Causal factors occur before the behavior in temporal precedence.
• Validity and biases, confidence in responses, and evidence quality are checked in research.
• Psychology focuses on intangible phenomena lacking definitive answers.

Basic vs. Applied Research

• Basic research addresses fundamental questions about the nature of behavior to increase understanding with unclear purpose.
• Applied research addresses practical real-world problems to find solutions, answer specific questions such as comparing teaching methods.

Research Starting Points

• Research designs can stem from testing assumptions, observing personal interests, addressing practical problems for solutions, or using theories to organize information.
• Past research can identify gaps and new avenues to explore using library databases or Google Scholar.
• Internet wikis provide general information but may lack empirical quality.
• A hypothesis is a statement about a potential phenomenon requiring future evidence based on existing theory, to support or refute it.
• Exploratory hypotheses develop clear hypotheses with logical explanations.

Scientific Study of Behavior - Steps

• Investigating human behavior involves asking questions, identifying existing knowledge, generating hypotheses, collecting and analyzing data, interpreting results, and disseminating findings through peer review and publication.

Research Proposal

• Research proposals identify novel research questions and hypotheses by reviewing existing research to identify gaps in knowledge and justify the research question and propose a study with methods, participants, sampling, procedure, measure, and design.
• It is similar to the introduction and methods sections of empirical articles.

Research Fundamentals - Variables

• A variable is any event, situation, behavior, or trait that can vary.
• The variable of interest is the one the researcher focuses on.
• Situational variables are features of the environment, that the participant is exposed to, while response variables are reactions elicited.
• Participant variables are pre-existing traits.
• Operational definitions specify the operations and techniques used to represent variables with measurement, predictions, and research designs, which can have multiple definitions.

Basic Research Designs

• Non-experimental designs, also known as correlational methods, determine relationships among the variables of interest and measure and observe variables without intervention to see if they change together.
• Non-experimental designs establish trends, describe behavior, and predict future behaviors, but have the disadvantage of the direction of causal influence.
• Experimental designs determine causal influences by manipulating at least one variable (independent) is measured and observed, where different levels of IV are different experimental conditions.
• The Dependent variables, that is affected by IV changes and are measured by research.
• The choice of design depends on the hypothesis, operational definitions, and the conclusions made.
• Experimental designs answer questions about the causes of behavior and have high internal validity and more experimental control but sometimes lack external validity and can be ethically impossible.

Non-Experimental vs. Experimental Designs

• Non-experimental designs determine relationships between variables while experimental designs determine causal influence.
• In non-experimental methods, no variables are manipulated, with all measured and sufficient for describing questions but cannot determine causality and have weak internal validity.
• In experimental designs, at least one variable (IV) is manipulated and one (DV) is measured, which can confidently determine alternate explanations.

Causal Influences

• Causal influences need temporal precedence (IV before DV), the IV and DV need to covary, and alternative explanations must be eliminated.
• All three criteria are met when there is high internal validity and the degree to which the experiment allows for causal conclusions.
• Cofounding variables can influence changes in IV and DV so experimental control is needed, participants must be treated the same and control individual differences.

Sampling and Measure

• Population refers to the entire group of interest, while a sample is a smaller subgroup that completes the study.
• Representative samples should be similar to the population for generalizable conclusions.

Sampling Probability Techniques

• Probability sampling gives all population members a known chance of selection.
• Simple random sampling selects participants randomly to create equal probability.
• Stratified random sampling divides the population into subgroups and selects proportionally randomly from each.
• Cluster sampling identifies population clusters and samples everyone within those clusters.
• Advantages of probability sampling are high representativeness, while disadvantages include resource intensity and difficulty accessing subgroups.
• In Cluster sampling, a population is divided into groups "clusters", and then randomly selects which clusters to sample.

Non-Probability Techniques

• Non-probability sampling does not ensure accurate representation.
• Convenience sampling selects participants based on availability.
• Purposive sampling selects participants based on availability and specific criteria.
• Quota sampling selects participants based on availability while proportionally representing subgroups.
• Non probability has advantages of inexpensive, less time, and convenient, and its disadvantage trades for having a less generalizable representative sample.

Measurements

• Self-reports involve individuals responding to their thoughts, attitudes, and beliefs.
• Self-reports include answering questions on a scale but might rely on the assumption that the participants are accurately honest.
• Behavioral measures code direct observations, occurrences of specific behaviors.
• Physiological recording measures body responses.
• Advantage to behavioral measures is the participant is unaware of how they act, but it can be costly and timely consuming and subject to researcher subjectivity
• The advantages to physiological responses are not subjective and observations occur outside of a individual conscious awareness, but are expensive.
• Measurement choice depends on operational definitions, research hypotheses.

Measurement Quality

• Quality in Measurement depends on cost, reliability, validity, and reactivity.
• Reliability refers to the consistency of produced scores. High reliability means the measure is stable.
• Internal consistency involves being consistently measured across items, testing reliability can be estimated using Cronbach's alpha.
• Test-retest reliability checks consistency across time; Inter-rater reliability assesses consistency across different raters.
• Validity assesses whether a measure accurately measures the intended behavior or construct, which has high construct validity.
• Reactivity refers to how much the measurement process alters participant behaviors. Nonreactive measures do not alter behavior. Measuring blood pressure and Self-reports can be reactive.

Ethics

• Research ethics prioritizes the well-being of participants during research, which is reviewed by a government board.
• Informed consent ensures that participants are aware of essential aspects of a study that may influence their decision to participate, inc risks.
• Observational research and special populations are exceptions to the fully informed requirement.
• Research may acquire deception but has to be debriefed upon competition of the assignment.
• Benefits of potential participation should outweigh risk and should maximize scientific validity.
• Inclusion and exclusion criteria must be justified for justice and the criteria should maximize validity and potential benefits.
• WEIRD: western, educated, industrialized, rich, demographic. Ensure fair access to all level risks.

Ethics Boards

• Ethical reviews require researchers to consider ethical and legal aspects, ensure participant well-being, and avoid scientific misconduct such as data fraud or unethical analysis to minimize threats to statistical results and to increase accuracy of transparency.
• Ethics boards have review processes to guide research and must undergo analysis to adhere and make sure any research is ethical.

Observational Methods

• The 2 types of Quantitative and Qualitative data: one results in numeric and data analyzed with frequency and intensity.
• Qualitative results in numerical data that's analyzed patterns and themes like test based material.
• Passive: videos or observations/ Non passive: images.
• Naturalistic observations involve systematic observations in natural settings and be qualitative or quantitative with 2 methods like participant or non, high external validity rich informed, it is the only way you will be able to approach but has limitations like lack of control, time constraint, observer bias and reactivity.
• Structured observation has codes in a lab setting depending on behavior that you identify, good external validity but reactivity.
• Case studies researcher gives detailed explanation of behavior from single individual; this is often qualitative, common for studying low frequency phenomena and offers a rich detailed description.

Self-Report

• Self-reports are data by asking behavior description from participants but relies on truth and attitudes and views , can be experimental or non.
• Questionnaires are sets of answered questions by participants towards attitudes and behavior; these are quick and information; close is a limited choice of responses.
• Relying on the truthful answers is issues for responses, if social bias exists or if there are fraudulent or misunderstandings, use correct wording to have reliable and use existing and appropriate choices.

Asking Good Questions

• To ask a "good" question, you have to make sure your question asks about one concept in a way that a question is framed negatively with not phrases, emotional language or jargon.
• Face-to-face interviews: Interviewer & participant is in person , open and small sample size.
• Telephone interviews/ Quantitative & Qualitative: Traditionally done by phone but can zoom or skype with big sample sizes.
• Focus groups/ Qualitative: Interview 6-8 participants code interactions rich data and costly.

Experimental designs

• You have to make sure that the social desirability look into good and characteristics of the interviewer and the experimenter has expectancy rate.
• Naturalistic is better than structured due to ethics , bias minimized but if there has to be cause and effect, structured is better if there is any cofounding variable.
• Basics of experimental: Design is to find any causes, the measuring if things change by the IV and validity with alternative plans.

Research Design

• Must have IV and DV. Need Operational definition and measurement and to look at the reliability, validity, and reactivity.
• The experimental design is divided in two parts to reduce any variables. IV is operationalized to at least have 2 conditions:
• In Between subject, there are different sets per level, starts as whole. Advantage one is that is easy to manipulate to get exposure.
• Within subject gets both all IV sets and has good identity to all and can compare to another, but this is threatened by issues coming from repeated measure.
• In order effects of preforming something and doing task again due to boredoms and is more accurate if can have two sets of people.

Experimental Considerations

• You will be able to give accurate effects that aren't accidental with only a fewer set participants.
• However its ethically can guess change behavior and cant give multiple level IV
• The things to use in which design should you use?: resources and external validity.
• Quasi experimental is like real, just has lack of feature if causes were involved: you don't know where to infer, no causal with history or natural change , groups by age, groups divided.
• With these it cannot be that control and comparison or where its unethical for random assignment.

Major Categories

• Cannot apply some techniques due to limits due to the pretest, its better to add is better to have more info on groups to compare but is not completely valid.
• Complex experiment is a mix with an IV of more than 2 levels or with another IV.
• Factorial are high designs each level means what to do based on factor.
• This is to ask amount of reactions to one and another?
• Main is what happens where to compare whats changed, and whats the interaction effect

What Interactions Can Be

Factoral research can be both, fully with in and mixed with 1 within, there has to useful or explained.

• Statistical, the study can be given a number with responses. There are to data collected for quantitative or qualitative. The research has to test hypothesis. If its to high a scale depends with the participant and what it could become.