PBSI Final Methods and Assessment Notes & Resources

S-Data (Self)

Self report questions

I-Data (Informant)

Information given by other people, such as friends and family

B-Data

Behavioral observations in a lab setting

L-Data

Verifiable information such as, education records, employment status, and marital status

Pros of S-Data

• Access to info

  • Ease of use

Cons of S-data

• Bias and error

Pros of I-Data

• Real world

• More objectice

Cons of I-Data

• Lack of access

• Bias

Pros of B-data

More objective

Cons of B-Data

• Interpretation

• Logistics

  • Ecological validity

Pros of L-Data

Verifiable evidence

Cons of L-Data

Multiple explenations

Mixed Data

Using a mix of data types

Online data collections forms

• Google forms

• Cloud research

  • Survey monkey

Pros and cons of data collection platforms

Pros: Efficiency, volume of information

Cons: Security and attention

Projective test

Respondents may be unaware of inner processes

• Rorshach inkblot test

• Draw-a-person test

  • Thematic Appreception Test

Pros fo Projective test

• Breaking ice

  • Access less conscious information

Cons of Projective test

• Subjectivity

• Less validity

  • evidence

Self report assessment

• The most common way to study personality

• Personal insight required

• Various questions format

  • True/False

  • Incremental agreement scale (Likert Scale)

• Social Limitations

  • Socially desirable responding

  • Acquiescence response set

Likert Scale

1-5, 1 is strongly disagree to 5 strongly agree `

Marlowe-Crowne Social Desirability Scale

• Measures tendency to engage in self-bias

  • Statistical control in assessment `

Types of research designs

• Case studies

• Correlational designs

  • experimental designs

Case Studies

• In depth study of an individual

• Holistic approach

• Unique, and sometimes too unique to apply to the population

Correlational Designs

• Examines relationship between response to two, or more variables as they naturally occur, typically simultaneously

• No manipulation

• Good for establishing patterns

• commonly used

Experimental designas

Systematically compares groups to determine difference in a response variable as a result of an independent variable

• Random assigment

• Control/comparison groups

• Casual conclusion s

• Not always possible

• Ecological validity concerns

Research practicer

1. Ethics

2. Open science and preregistration

3. Generalizability

4. Use quality measures

Research practicer: Ethical research

• Informed consent

• Belmont report

  • Autonomy beneficence

  • justice

• Institutional review boards

• Institutional animal care and use committee

Research practice: Open science and preregistration

• Transparency and honest as scientific foundation

• Avoid plagiiarism and fabrication of data

• Report data completely

• Fully describe all aspects of all studies

• Report studies that failed and succeeded

• Freely share data

Open science framework

• Tool that scientist use to track their progress to ensure their work is verifiable and honest

Generalizability

The degree to which you can apply the results of your study to a broader contex

Multiple methods and quality measures

• Replication > confidence

  • Across studies, labs, and methods

• Quality data > confidence

Meta-analysis

Synthesis of results accross many studies on same topic

Statistics Primer

1. Descriptive Statistcs

2. Statistical significance

3. Effect sizes and power

   1. Correlations

Descriptive Statistics

Describe properties of a dataset such as similarities and differences

• Central Tendencies

• Variability

Many variables follow a common shappe, the normal distrubution

Descriptive Statistics: Central Tendency

• Mean

• Median

  • Mode

Descriptive statistics: Variability

Range

Standard deviation

Statistical Significance: Null- Hypothesis significance testing

Determins the chance of getting the result if nothing were really going on

Statistical Significance: P-Value

Probability of obtaining a result if there is no difference between groups or no relationship between variables

Statistical Significance

• *= Alpha level

• NHST

• P value

Many people are critical of this method

Effect Sizes & Power

• Addresses practical significance

• Measures of strengths of relationship between variables

  • Correlation coefficient, r

  • Cohen’a

• Power

Effect Sizes & Power: Power

Probability of detecting a significant effect in your study, assuming that it does exist

• INcreass with effect size

Correlations

Effect size dipicting magnitude of lenear relationship between two numerical values

Psychometric properties

Ideally, a strong measure will have these properties

• Reliability

• Validity

Psychometric properties: Reliability

Consistency in measurement

• Internal consistency

• Test-retest

  • Interrater

Psychometric properties : Validity

Measuring intended quality

• Face

• Predictive

• Convergent

• Discriminant/divergent

Scale Development

Requires theory, collaboration and statistical analysis

• Factor analysis

• 4 steps

Scale Development: Factor Analysis

A statistical technique that identifies groups of things that seem to have something in common

Scale Development: Step 1

Generate a long list of objective items

Scale Development: Step 2

Administer these items to a large of people

Scale Development: Step 3

Analyze with factor analysis

Scale Development: Step 4

Consider what the items that group together have in common and name the factor

Barnum Effect

Tendency to believe vague, especially positive statements about oneself