Double Blind

Mean

"average" score...Interval & ratio data. Best for continuous data. If skewed, mode or median will be more accurate.

Experimental designs - Pros

+ Explores the CAUSAL relationship between variables + Random assignment for independent variable + Variables not being examined may be held constant (using control group)

Experimental designs - Cons

- Experimenter may give subtle cues to participants on how to respond...Expectancy is an issue, but solvable with precautions of standardized procedure & double blind procedure.

Data analysis includes:

- Open coding - Axial coding - Selective coding

Interviewing Types

1) Informal conversational interviews 2) Guided approach *Best for children, with their cognitive limitations, linguistic limitations, and attention span 3) Standardized open-ended interview

Qualitative Research Characteristics

1. Depth Rather than breadth 2. Learn about "how and why" people behave etc. 3. Often Exploratory 4. Often inductive 5. Continual shaping and reshaping of the data 6. Data is in the form of words, images 7. May be a follow-up of previous research where a quantitative researcher has reached an impasse in explaining or understanding.

Trustworthiness of Qualitative Data

1. Persistent Observation 2. Prolonged Engagement - data gathering 3. Thick, Rich, In-depth Description 4. Triangulation of methods, sources, investigators 5. Negative Case Analysis 6. Audit Trail 7. Conceptual Saturation 8. Member Checks 9. Peer Debriefing 10. Explicit Documentation of data collection methods, analysis, and field decisions that altered any strategies or focus

Ethics

1. Tuskegee Syphilis Study (1930-1972) 2. Belmont report (1974)

Triangulation of Sources

3 different types of sources

Elements of a Research Paper

Abstract Introduction Methods Results Discussion

Convenience

All are convenience, but specifically a "convenience sample" makes no attempt to randomize population. You can tell it apart from the other non-probability samples because it lacks the criteria, snowballing, and subgroups of the other types.

Inter-item consistency (reliability)

Asks the same question 500 times. Kronbachs Alpha. How unified the items are in a test.

IRB Process

At least 5 members A person employed by the institution A person NOT employed by the institution A scientist A non scientist One other

Belmont Report (1974)

Beneficence - No harm! Justice - make sure no group is "picked on" or benefited in particular Respect for person *(IRB came later)

Double Blind

Both people giving and getting test don't know what procedure they are getting.

Content Validity

Breadth of items in the measure

Inter-rater (reliability)

Consistency of results when the same test is given by different researchers. uses Cohen's Cappa

Methods

Design Sample Measures & Operational definitions Procedures

Post Hoc

Doing research on something that couldn't be ethically tested.

Tuskegee Syphilis Study (1930-1972)

Effects on black people vs. white people...When the real antidote for syphilis came out in 1950's, it was denied being given to the blacks...People went blind and many died.

deductive

General theory -> specific data

Test-re-test (reliability)

How stable a test is over time.

Introduction

Hypothesis Review of literature Conceptual definitions (i.e. this is what happiness means) Contribution

Directional

If there is more infidelity, there will be less happiness. There is a relationship between infidelity and marital happiness.

7 Elements of Experimental Design

Independent Variable Dependent Variable Pre-test Post-test Experimental Group Control Group Random Assignment

Discussion

Interpretation of results, Implications & Limitations

central tendency

Mean Median Mode

General source

News or news paper, or a novel

limitations of Correlational data.

No proof of causation, only correlation (makes sense with the name).

Different types of measurements - NORI

Nominal (categorical) Ordinal (scale from 1-10) Ratio *0/zero matters (ex:How many times did you brush your teeth today?...0 is a possible answer and an issue) Interval (ex: waist size or temperature) *No meaning to zero/zero is arbitrary

One group pretest-posttest

O - Assessment X - Treatment O - Assessment

The research question is: "What important themes do young adults identify from the April 2011 General Conference addresses?

Open coding - identify the major themes for each participant. Some participants may have more than one theme coded (double coding). Talk with several others in the class to identify if you have the same themes (triangulation of investigators) List the major themes you have come up with Axial coding - look at your major themes - are there any themes that could be collapsed into a broader category? Any sub-categories? Selective coding - what quote best represents each of your themes?

primary

Peer reviewed ACTUAL journal articles detailing research findings

Non -Probability Sampling

Purposive Snowball Quota Convenience

Classical Experimental Design

R - Random Assignment O - Assessment X - Treatment O - Assessment

Post-Test Only Control Group

R - Random Assignment X - Treatment O - Assessment

Axial Coding

Refining your categories

Selective Coding

Representing the whole - look selectively for cases that illustrate each theme. Identify best quotations. Identify how themes best fit together. (ex. Pick a favorite from your ten groups. he has a great quote that I am going to use ) *It's always non-probability because you have a broad question.

Probability Sampling

SRS (Simple Random Sample) Cluster Multi-stage Cluster Systematic Stratified

Convergent Validity

Similar enough - *correlation about .8 or more

Inductive

Specific data -> general theory

Abstract

Summary

Open Coding

The initial "pass" through the data. Locate initial themes and name (code) them. Creates many new themes.

Quantitative Research

The use of sampling techniques (such as consumer surveys) whose findings may be expressed numerically, and are amenable to mathematical manipulation enabling the researcher to estimate future events or quantities.

Non-Equivilant Control Group

There are 2 groups. One group receives a treatment, and the other does not, but both are tested them same way. X O, and O

Non-Directional

There isn't a relationship between infidelity and marital happiness

Content Analysis

Used to identify content (messages, themes) in various types of materials. Identify the research question. Identify the unit of analysis.

Confounding Variable

When variables compete to explain the affects found in a study

One-shot case study

X- Treatment O - Assessment

Purposes of Critical Research: To change the world through...

a) Empowerment - especially for less powerful groups (ie. women, racial minorities, people with disabilities, etc). b) Enlightenment - uncovering myths; exposing false ideologies. c) Emancipation - freeing people from false ideologies. Helping people change the world for themselves

Internal Validity

accuracy in concluding that the outcome of an experiment is due to the independent variable

Discriminative Validity

are you different enough - *correlation about .2 or less

Median

arranged scores lowest to highest & the median is in the middle. Best for rank order/ordinal data or unevenly distributed continuous data (possibly ratio or interval)

Triangulation of Method

asking questions

Confirmability (objectivity)

can the findings of the study be confirmed by someone else?

Purposive

choose participants based upon aims and purposes of the study

Conceptual Saturation

collect data until no new categories are generated.

Results

data

Credibility

demonstrates that the inquiry was conducted in such a manner to ensure that the participants were accurately identified and described.

Systematic

every Kth element of list is selected

Qualitative Research

explores processes that underlie human behavior

Face Validity

face to face, on the same level as fellow researchers

Prolonged engagement

gathering data over time

SRS (Simple Random Sample)

generalizable, but takes time & a large sample

Quota

get a certain number *used for representing the important subgroups in a population.

Persistent Observation

going back and re-interviewing recurring observation in participants.

Criterion Validity

how does the measure relate to already known standards

Reliability

how stable or consistent the measure is

Concurrent Validity

how well a test measures present performance

Predictive Validity

how well it predicts future performance (correlation coefficient)

Correlation coefficient (reliability)

index of strength between 2 variables. Difference between test A and B.

Researcher

is an important part of the qualitative process - he/she is the primary source.

Validity

it is measuring what you say it is

Cluster

list of clusters, random sample (ex: classroom) Sampling a cluster means they are all together in some form, like a club or quorum.

secondary sources

most books, Time magazine (second summary)

Explicit Documentation

of data collection methods & analysis at each stage

Construct Validity

operational definition accurately assesses theoretical concepts or results of the test are related to an underlying psychological construct.

Snowball

participants nominate participants

Stratified

population divided into strata and random selection made (ex: male & female strata, or gender strata)

Peer Debriefing

present analyses and conceptual abstractions of the data to other qualitative researchers (to explore inquirer bias and clarify meanings and the basis of interpretation)

Member Checks

present analysis of the data to informants for their confirmation or revision

Persistant Observation

recurring observation of participants

Directional Hypothesis

reflects a difference between groups and the direction of the difference is specified (if-then)

Qualitative Research

research that examines phenomena within the cultural and social context in which it takes place

Negative Case Analysis

search for and explain phenomena that do not fit. Revise hypotheses until they account for all known cases.

Social Citation Index

seeing how many times an article is cited

PyschInfo (EBSCO)

seeing specifics of articles, secondary source.

Multi-Stage Cluster

several levels of clusters, then sub-sampled

Thick-Rich in depth description

talk about the details

Threats to internal validity

the accuracy in concluding that the outcome of an experiment is due to the independent variable.

External Validity

the extent to which results of an experiment can be Generalized.

Mode

the score that appears most often. Best for categorical data. *Can be 2 or more.

Audit Trail

theoretical memos including detailed descriptions of how you got from the data to the conclusions.