You are on page 1of 11

Lecture 2: What is scientific process?

Characteristics of Scientific research

*empirical: Based on evidence

*logical: follows the rule of logic

*Systematic: there is an inherent order of conducting research

*objective: independent of subjected elements, presences, biases.

*Falsifiable: scientific theories must always be stated in a way that the predictions derived
from them can potentially be shown to be false

*generalizable: social research seeks to understand aggregate regularities

*It is a process: These is no absolute truth in science, our conclusions are tentative; should be
modified by new knowledge

Lecture 3: Types of research:

Goals of behavioral science:

*Description: systematic observation. Addresses the amounts of a single given characteristic.

*Prediction: Research for discovering regularities, relations between variables.

** 2 or more variables

**Series of coordinated observations.

**Strength and direction of relation

*Finding causes and Explanation: Research for explaining situation and events.

Identifying causality:

*temporal precedence

*co-variation of cause and effect

*elimination of alternative explanations


Technique of data collection

Quantitative (Nicel) research uses

*Numerical data

*Explicit and measurable

*statistical analysis

*Potential of losing the richness of meaning

Qualitative (Nitel) research uses

*Non-numerical data

*Meaningful within their context

*Able to recognize the spectrum of a concept.

*Potential of ambiguity

Lecture 4: Research question

*Hypotheses: Is variability in one variable related in a systematic fashion to variability in


other variables.

Two types of variance:

*Systematic variance: Behavior that is related in an orderly, predictable fashion to other


variables

*Error variance: Portion of total variance that is unrelated to the other variables.

Research process:

1: Research ideas: What’s your question

2: Convert idea to hypothesis

3: define and measure variable

4: Identify participants or subjects

5: Select a research strategy

6: Select a research design


7: Conduct the study

8: evaluate the data

9: report the result

10: Refine or reformulate your research idea.

11:2: Convert idea into hypothesis

Lecture 5: ethics

Responsibilities to participants/subjects:

*Treat human research participants with respect and in a way that maintains their rights, and
dignity.

*Care for the welfare of animals when they are subjects

*Be hones in the treatment of data

Major ethical concerns:

1 Harm vs benefit

Loss of time and resources, physical and psychological and physical harm, social or political
harm vs educational benefits, new skill or treatment, material benefits, personal satisfaction.

2. Voluntary participation/ informed consent

*describes what subjects can expect in study

*time commitment

*potential risks or discomforts

*benefits

*participation is voluntary and subjects are free to withdraw

3. Privacy and confidentiality

4. Deception

5. Debriefing
*full description of the true purpose of the study- including the deception, presented after the
experiment.

Responsibilities to the discipline

Types of scientific misconduct:

*Plagiarism

*Fabrication

*Falsification: overstated/ over-interpreted results, manipulation of data equipment.

*Confirmation bias

Institutional approval

*Institutional review board (irb): Each institution is required to establish a committee, that
reviews and approves the procedure of the study

Lecture 6: Hypothesis formulation

Research Hypothesis: convert your specific research idea into a hypothesis

*I want to research the effect of secrets on perception too general

*I want to research the effect of recalling an important secret on students’ distance perception
better but not measurable and has no direction

*Students that recall an important personal secret will verbally overestimate the determined
distance compared to students who recall a memory. (testable, measurable, specific,
directional)

Variable is the opposite of constant/ attribute. Some variables are intangible (not directly
observable, like racism)

Hypothesis: x (first variable) >(effect and direction of effect) y (second variable)

Elements of a hypothesis:

1 variables and their definitions (conceptual and operational)

2 relationship between variables


3 theoretical justification: why do we assume the relation between variables to be in a specific
way.

Types of Hypothesis:

Research/ alternative/ experimental hypothesis: Hypothesis that you are testing

Null hypothesis: The logical opposite of your alternative hypothesis

 A statement about what you would expect if there was no relationship between the
variables in your study.

One-tailed Hypothesis (directional): Ex. Female’s weekly alcohol consumption will be less
than males’

Two-tailed Hypothesis (non-directional): there will be a difference between males and


females’ in their weekly alcohol consumption.

Priori Hypothesis: Hypotheses should be formulated before the actual study

Lecture 7: Hypothesis and theories

Deductive model of research: theory > hypothesis > operationalization > observation >
analysis > conclusions

Inductive model of research: observation > organization > general statement > theory

A scientific theory is: A collection of general principles that outline how a particular
phenomenon works.

*A logically organized set of propositions that serves to

**Define events

**Describe relationships among these events

** Explain the occurrence of these events

Scientific theories are

*Grounded in data and previous hypotheses, but

** is more than a simple hypothesis

** falsifiable through hypotheses


*modified when necessary

*should be parsimonious (simplest and shortest statement that cover the facts)

Micro theories: local theories, relating to a specific area, group of people or aspect of the
social world.

Meso theories: Middle level theories relating to social phenome usually found, such as
organizations, institutions, community, family.

Macro theories: Theories that attempt to cover all aspects of the social or natural world in
general terms.

Lecture 8: Relationship between variables

Types of relationships:

 Positive vs negative relationships


 Strength of the relationship: The extent to which two variables are related. (correlation
coefficient)
 Symmetrical (correlation/ prediction)vs asymmetrical relationships
(causation/explanation)
 Linear vs curvilinear relationships
 Spurious relationships (3rd variable) (confounding variable effects both variables)
 Intervening (mediating) relationships (we see a>b but a>c>b )
 Interacting (moderating) relationships (relationship of two variables change for
different categories of 3. variable)

Lecture 9: operational variables and scales of measurement

Types of variables: variables can vary in type, kind, or categories of a phenomenon


(categorical) or in amount or degree of a phenomenon (quantitative).

*independent/causal variable: comes first and influences or predicts other variable.


(Antecedent)

*dependent/outcome variable: comes second (consequence)

*Third variable; mediator, moderator, extraneous (confunding).


*Subject variables: variables that can’t be manipulated. (gender, age)

Conceptual definition: Specifying the exact meanings of all the concepts studied.
(conceptualization = meaning )

*Includes the process of identifying dimensions and indicators of our constructs.

** A dimension is a specific aspect of a concept.

** An indicator is an observation that we choose to consider as a reflection of a variable we


wish to study

Operational definition: Definition of the variable in terms of the operations or techniques the
researcher uses to measure or manipulate it.

Ex. Intelligence

*conceptual definition: A person’s general ability to adopt to his or her environment

* operational definition: A score in the WAIS test

Operationalization:

Process of moving from abstract to concrete.

Steps of Operationalization:

1: Determine the variables and their attributes: Values must be exhaustive and mutually
exclusive.

2: Determine the unit of analysis: What or whom to study

3: Determine the level of measurement

*Quantitative variables can be discrete (no fractions) or continuous (contains fractions)

*nominal scale: categories but no order to the categories. All qualitative measurement is
nominal.

*Ordinal Scale: Ranked or ordered categories

*Interval scale: Categories are approximately equally spaced or ordered. (Ex. Temperature in
Celsius)
*Ratio scale: Equal intervals with a true/ absolute zero that indicates the absence of quantity
(temperature in kelvin)

4: determine the modality of the measurement

*Behavioral measurements: Overt behaviors that can be observed

*Physiological measures: Brain imaging

*Self-report measures: Surveys, interviews

*Use already existing statistics

5: Determine whether to use single or multiple indicators (tangible vs abstract variables):


Many concepts have multiple dimensions with multiple indicators.

*Multiple indicators:

**Scale: composite, multiple measures about a particular aspect of a concept.

**Index: Composite measure of multiple dimensions of a particular concept.

Lecture 10: Research design, external and internal validity

1 One or many variables?

Univariate: Questions with single variable.

Multivariate: Questions with relationship between two or more variables.

2 Relationship or casual explanation:

Experimental (causal) vs non-experimental (correlational) methods:

*Experimental: involves direct manipulation or control of variables

*non-experimental: relationships between variables are studied by making observations/


measurements as the variables occur naturally

Establishing causality: Requires one to overcome

*directionality

*third variable problems

*achieved by manipulation, experimental control, or randomization


3. Understanding few in detail or many in less detail

Qualitative data usually has more details but fewer cases.

Quantitative data vice versa.

Research validity:

The criterion of quality of research design.

*Statistical Conclusion validity: Appropriateness of statistics

*Construct (measurement) validity: appropriateness of operational definition (measurement


tools)

*Internal validity: confounding variable, participant reactivity (change in behavior observed


in participants caused by social desirability), experimenter bias, selection bias…

*External validity: non-generalizable findings

Problems with multiple (longitudinal) measurements:

Social and environmental events can interfere with the experiment

Maturation: Subjects may mature or become disinterested

Instrumentation: Measurements may not be comparable across different administrations.

Attrition/experimental mortality: Subjects give up or drop out.

Statistical regression: Participants chosen due to extreme scores, behavior may return to
normal range when tested again.

Internal validity:

Ability to draw conclusions about causal relationships from the result of a study.

External validity:

Ability to generalize research findings to the population of interest.

Types of external validity:

*Population validity: Accessible population, target population

*Ecological validity: different settings


*Temporal validity: Across time

*Treatment variation validity: variations of the treatment

*Outcome validity: different but related dependent variables

Real life situations: high external validity, low internal validity

Lab settings: vice-versa

Theoretical position is more concerned with internal validity

Applied research is more concerned with external validity

Lecture 11: Measurement Reliability and Validity

Measurement experts are called “psychometricians”

Measurement Reliability:

Refers to the consistency or stability of a measure or behabior.

Types of measurement reliability:

*Stability:

**Test-retest

**Equivalent/alternate/parallel forms: Measure the correlation between alternative


instruments, designed to be as equivalent as possible given to the same group of subjects.

*Internal consistency

**Coefficient a (Cronbach’s alpha): calculates simultaneous correlations and comes up with a


summary reliability score. (Consistent if alpha >0.70)

**Split-half reliability: randomly selected half agrees with the other half.

*Inter-rater reliability: Assessments in which there are more than one observers/raters. (extent
to which raters agree in their observations)

Measurement of validity:

Measurement/construct validity

*Face validity: At the surface level, dues it look as if the measure is testing the construct.
*content validity: Refers to the extent to which a measure represents all aspects of a given
construct.

*criterion validity: Obtained by relating your measurement scores with one/more relevant and
known criteria. (criterion is a standard you want to predict on the basis of your measurement)

**Predictive validity: How well the measurement of a construct accurately predicts future.

**Concurrent validity: How well the measurement of a construct accurately predicts


behaviors and outcomes measured at the same time

*Convergent validity: How well the scores received from the new method correspond with
scores obtained from other measures of the same construct

*Discriminant (divergent) validity: How different the scores are from scores obtained from
other similar methods designed to measure theoretically different constructs.

Range effect: Adequate distribution of responses to be able to detect difference at one


direction. (ceiling or floor effect)

You might also like