• Shuffle
    Toggle On
    Toggle Off
  • Alphabetize
    Toggle On
    Toggle Off
  • Front First
    Toggle On
    Toggle Off
  • Both Sides
    Toggle On
    Toggle Off
  • Read
    Toggle On
    Toggle Off
Reading...
Front

Card Range To Study

through

image

Play button

image

Play button

image

Progress

1/83

Click to flip

Use LEFT and RIGHT arrow keys to navigate between flashcards;

Use UP and DOWN arrow keys to flip the card;

H to show hint;

A reads text to speech;

83 Cards in this Set

  • Front
  • Back
  • 3rd side (hint)
What is the scientific method?
Method for answering questions (inherently flawed).
Only good for answering questions about the physical world.
Group effort- what's important is the knowledge derived, not the people.
Describes, predicts, explains and determines causes of behavior.
Assumptions of Science
(LUDE)
Lawful- orderly, not random, predictable sequence of cause and effects.
Understandable- with effort, one can understand the deterministic and lawful nature of the subject matter.
Deterministic- subject matter is determined by other "natural factors"; not due to free will.
Empirical- knowledge derived from the senses (soft science based on hypothetical constructs which are made sense of by operational definitions and what we infer from them).
Attitudes of Science
(OSU)
Open-minded- there may be various interpretations of data.
Skeptical- all theories and beliefs derived from science will eventually be proven incorrect; the good theories are the ones that withstand being proven incorrect the longest.
Uncertain- want to remain uncertain about the accuracy of scientific conclusions due to the inference involved.
Epistemiology
Scientific approaches to knowledge:
Logical positivism
Humanistic perspective
Social constructionism
Logical positivism
Use of tightly contolled experiments, logical analysis of the data, knowledge for knowledge's sake, "distance" between experimenter and subject matter (i.e. removal of bias).
Humanistic perspective
Science should serve the greater good, studies should occur in the natural environment, understanding is derived from intuition and empathy.
Social constructionism
All knowledge is a product of socio-historical processes, knowledge represents the scientist's reality, knowledge is a product of "culture" (used in of broad sense), science is one way to construct a knowledge base.
Hypothesis
An educated guess about a relationship between 2 variables.
Variable- something that varies; must have at least two levels.
Relation- one variable systematically changes with another.
Criteria for hypotheses (Take Front Row Pics)
Testable- have to have an empirical test for your hypothesis.
Falsifiable- has to be able to be proven incorrect.
Rational- has to fit with existing knowledge (ie link to literature).
Parsimonious- simple, concise with few assumptions.
Dependent Variable (DV)
The "effect" in the cause-effect chain. The dependent variables are those that are observed to change in response to the independent variables.
Independent Variable (IV)
The "cause" in the cause-effect chain. The independent variables are those that are deliberately manipulated to invoke a change in the dependent variables.
Descriptive Statistics
Used to describe a sample (mean, median, mode, range, domain, standard deviation, variance)
Inferential Statistics
Tell you that the probability that you see something in the sample that will occur in the population (t-tests, ANOVA, MANOVA, regression)
MAXMINCON Principle
MAXimize the variability in the DV that is associated with the variablitiy in the IV (easiest way is to use a strong IV)
MINimize the variability in the DV that is associated with random factors (use as large a sample as possible)
CONtrol the variability in the DV that is associated with variables other than the IV (ie. confounds/extranious variables)
What are the main 3 weaknesses in science?
Flaws in...
Evidence
Inference
Hypothesis testing
Flaws in Evidence
-quality of operational definitions used
-quality of the measurement used for the variables of interest
-science is a "snapshot" and only accurate for that point in time.
Flaws in Inference
-when the hypothetical construct becomes a "thing" (when it ceases to be a description)
-pseudoexplanations- "circular reasoning" that seems like an explanation when it isn't
Flaws in Hypothesis Testing
-confirmation or looking for evidence that supports your hypotheses, rather than refuting them
-failure to use replication
Population
The entire group about which the researcher is interested.
Sample
The subgroup of the population that participates in the study.
What are the big question and big problem in locating a sample?
Question: How well does the sample represent the problem?

Problem: Sample procurement- you have to pay attention to how you're getting your participants and whether they really represent the population you are attempting to study.
What is the big question in identifying the hypothetical construct?
How well does the hypothetical construct capture the psychological factor of interest?

**You must consult professionals (people or literature)
What are the big question and problem in identifying variables used to infer the hypothetical construct?
Question: How well do the existing variables capture the hypothetical construct?

Problem: How frequently a variable is used versus how good it is.

**A good variable is one that captures all the important aspects of the hypothetical construct.
How do we describe relations when testing hypoetheses? (Keywords: relations, strength, type)
Relation: systematic change in variable X with variable Y.

Strength: the consistency in change between X & Y.

Type: the direction in change in X compared to the direction in change in Y. (Positive = same direction; Negative = opposite direction)
What is the role of descriptive statistics?
-Summarize scores
-Describe relations between variables
-Predict an unknown variable from a known variable
What is the role of inferential statistics?
The confidence with which we can assume the relations between variables found in the sample are present in the population.

**We do this with our hypotheses: Null (no relation btwn variables) and Alternative/Research hypothesis)

**We are testing the validity of the null hypothesis.
What are the 3 main characteristics of an experimental design?
1. Manipulation of an IV (If you are not manipulating an IV, you do NOT have an experimental design!)
2. Random assigment to levels of the IV (conditions).
3. Control of known confounds.
What are ways to control confounds? (*5)
-Random assignment (does not make groups different but may make it difficult to see IV-DV relations)
-Experimental control (only select people with the known confound; however, this limits sample size)
-Statistical control (measure the confound and then statistically remove it; does does not limit sample size)
-Match participants on EVs (confounds)
-Blocking (build the EV into the study as an IV)
Quasi-experimental design
When you're lacking at least one of the three characteristics required for an experimental design.

**Most clinical research falls in this category.
Reliability and its 4 types
The consistency or repeatablity of your measures! It is marked by consistency over time and performing the same way in different situations.

1. Test-retest: simple correlation; stable over time, multiple measures.
2. Inter-rater: Cohen's Kappa; simple correlation; agreement between raters
3. Internal consistency (split-half): Cronbach's Alpha; average correlation between items; average correction between items and the score; degree to which the items measure the same underlying construct.
4. Alternate forms: very few tests have this.
Validity and its 5 types (CCIEE)
The extent to which a procedure measures what it's intended to measure. **Remember that validity is inferred, not measured! It refers to conclusions made from the measurement, not the measurement itself.
CCIEE: Content, Construct, Internal, External, Ecological
Content Validity
Degree to which measurements reflect the variables of interest.
Construct Validity and it's types
OR FACE validity

The degree to which measurements appear to measure the hypothetical construct.

Convergent: btwn 2 instruments that measure the same thing.
Discriminant: btwn 2 instruments that measure different things.

Criterion-related validity is also included here: how valid are the predictions you're making with the instrument.
*Predictive- predicting something in the future (ex. SAT/GRE)
*Concurrent- accuracy of score to situation (ex. depression)
Internal Validity
Mathematical relation reflects the relation between 2 variables of interest and only those variables.

-Low internal validity = the apparent relation reflects a relation with other variables (ie. confounds)
External Validity
Degree to which you can draw the correct inferences when generalizing beyond a study.

Ex. the problem of the undergraduate subject pool
Ecological Validity
Extent to which research can be generalized to common behaviors and natural situations.
**Courtroom example- a study may have external validity in that having a mock-jury in a classroom may approximate results from a real courtroom trial. But it doesn't have ecological validity in that the experiment does not in any way approximate a real courtroom trial, feel, or experience.

**You must strike a balance between internal validity and ecological validity.
Manifest Variable
The empirical (measurable) representation of the HC.

Ex. HC=depression, MV=score on BDI-2
What reliability coefficients and values are we striving for?
Always a correlation coefficient or r value (% of shared variance, ie overlap btwn variance).

**.80 is high enough, but greater than or equal to .85 is preferred.
What values are expected for validity coefficients?
Less than .30 is not acceptable, however you will hardly ever get over .60.
What is measurement error?
Unaccounted for error OR variation in the MV that is not associated with the HC. VALIDITY
What are 2 factors that can influence reliability?
1. Test length (fatigue)
2. Rating scale used (nominal, ordinal, interval, ratio). Ex. A 7-pt Likert would be more reliable than a 3-pt.
What is the relationship between reliability and validity?
A valid measure is reliable, but a reliable measure is not necessarily valid. Reliability limits validity.
What are the threats to internal validity? (*8)
1. History- events that occur during the study.
2. Selection- how participants are obtained; groups differ at the beginning of a study.
3. Statistical regression- when participants are selected based upon extreme scores or when pre- and post-testing is used, the scores tend to regress toward the mean.
4. Maturation- any kind of natural developmental change in the participants.
5. Pretesting- the experience of being assessed or doing a self-report measure prior to the start of a study.
6. Instrumentation- technology changes during the course of the study.
7. Attrition
8. Expectancies- of the experimenter and/or participants
9. Mortality
What are solutions to internal validity threats?
Random assignment
Matching- participants on certain characteristics
Within participant designs- where everyone in the study receives all levels of the IV
Statistical control
Blind and double blind procedures- combat expectancies
What is the difference between external and ecological validity
For a research study to possess ecological validity, the methods, materials and setting of the study must approximate the real-life situation that is under investigation. Unlike internal and external validity, ecological validity is not necessary to the overall validity of a study. You may have one without the other. Just because a study has external validity and is generalizable to the population does not mean the study was a miniature replica of a real-life situation.
How can sampling affect validity?
-Could affect both internal and external validity; ex. using volunteers vs. nonvolunteers (motivation)
-Sampling error- random error introduced from using a sample rather than a population.
-Sampling bias- error introduced because of the particular sampling procedure used; some will be random error and increasing your sample size will not help this.
List and define the 4 random sampling procedures. (SSSC)
1. Simple random sampling- everyone in the population has an equal chance of being in the study.
2. Systematic random sampling- the n is known before sampling begins and you stop once you reach the n for the group.
3. Stratified random sampling- strata (ex. according to gender, ethnicity,location, SES).
4. Cluster random sampling- similar to simple; treating a group, rather than an individual as the participant (ex. new teaching methods- classes are randomly sampled rather than individual students).
Define non-random sampling and discuss how it affects a study.
Non-random sampling: NOT every person or cluster has the same chance of being in the sample.
*You can have all stratified and cluster nonrandom.
*May affect IV-DV relation and therefore internal validity.
*May affect findings and therefore external validity.
List and define the 4 types of non-random sampling. (PSCS)
1. Purposive sampling- selecting a sample based upon characteristics.
2. Sampling for homo/heterogeneity- only want males/females, people vastly different from one another.
3. Convenience sampling- easy for the researcher to get.
4. Snowball sampling- re-recruiting from a previous sample, friends of friends.
What are the 4 scales of measurement? (NOIR)
Nominal- categories (ex. religion, gender)
Ordinal- ranking of categories (ex. military ranking)
Interval- equal distance between values on the variable without an absolute zero pt (ex. temperature, WAIS intelligence score)
Ratio- equal distance btwn values with an absolute zero pt. (annual income)
What are the 3 measures of central tendency?
Mean- average
Median- middle
Mode- most frequent
What are the 3 measures of variability?
Mean absolute deviation
Variance
Standard deviation

*Variability is a measure of how spread out a distribution is.
Explain variance and standard deviation.
The variance and the closely-related standard deviation are measures of how spread out a distribution is. In other words, they are measures of variability.

The standard deviation formula is very simple: it is the square root of the variance. It is the most commonly used measure of spread.

An important attribute of the standard deviation as a measure of spread is that if the mean and standard deviation of a normal distribution are known, it is possible to compute the percentile rank associated with any given score. In a normal distribution, about 68% of the scores are within one standard deviation of the mean and about 95% of the scores are within two standard deviations of the mean.

The standard deviation has proven to be an extremely useful measure of spread in part because it is mathematically tractable. Many formulas in inferential statistics use the standard deviation.
Explain z-distribution.
The z-distribution is the same thing as the standard normal distribution. A z-distribution is the same shape as the raw score distribution and a z-score describes the location and distance from the mean for a score. The mean is always 0 and the standard deviation is always 1.
What is the equation for a z-score?
z= (x-xbar)/SD
Central Limit Theorem
Regardless of the shape of the distribution of raw scores, the shape of the distribution of sample means will always be normal. Theoretically, the standard deviation of the sample then equals the standard deviation of the population so we can look at a sample and make inferences about the population.
The mean of sample distribution and the population should will always be equal.
What are 4 sources of research questions?
Theory, Practical Problem, Literature, Personal experience
What is the "right" research question.
Non-directional
Empirical
Narrow
Important
Grounded in existing literature
Hypothesis
A directional statement about the relation btwn 2 or more variables.
"It is hypothesized that there will be a relation btwn Construct A and Construct B such that the directional measure of A will be associated with the directional measure of B.
Null hypothesis
No relation exists btwn the variables.
Alternative/research hypothesis
A directional relationship exists btwn the variables.
What is the first and most important step in research project?
Literature Review
What are the goals of the lit review?
Immersion into existing literature
Identification of typical designs, populations, data collection methods
To tell your story (how you make sense of existing literature)
What are you looking for and where?
Look for:
Theory, hypotheses to be addressed, methodology, data analysis
Look in sources:
Primary- from the author(s) and peer reviewed.
Secondary- includes book chapters, summaries and lesser peer review
Tertiary- non-peer reviewed (NEVER USE)
What are the 4 components of statistical power?
Sample size is simply the number of people or units available to be studied.

Effect Size is simply the ability to detect an effect relative to the other factors that appear in your study.

Alpha level refers to the likelihood that what you observed is due to chance rather than your program.

Power is the likelihood that you will detect an effect from your program when it actually happens.
What does a "significant" result mean?
It means that the probability has reached the critical value and you reject the null hypothesis. It does NOT mean that you accept the alternative hypothesis.
What a does a "non-significant" result mean?
Failure to reject the null hypothesis. Do NOT say that you accept the null.
One-tailed vs. two tailed tests
directional vs. non-directional.
Type I Error vs. Type II Error
Type I: rejecting the null when it is true (alpha)

Type II: Failure to reject the null when it is false (beta)
t-test
The t-test assesses whether the means of two groups are statistically different from each other. This analysis is appropriate whenever you want to compare the means of two groups.

**The DV must be on interval or ratio scales because you are looking at mean differences.
The null would be that the group means are equal. The alternative would be that the group means are not equal.
What types of t-tests are there?
One-level= one-sample t-test
Two-levels= independent or dependent samples t-test
Why use a one-sample t-test?
Some times you have to decide if a sample mean is different from a hypothesized population mean.
Why use an independent samples t-test?
When you are comparing the means of two independent groups. For example, control group versus experimental group.
Why use a dependent samples t-test?
When you are looking at the difference between means within-groups. For example, before & after scores (pre vs. post test). You are actually looking at the difference between each participants scores.
What is the difference between a t-test and a simple ANOVA for the comparison of two means?
NOTHING! They are the same thing. F = t^2

T-tests are just a special case of ANOVA.
What is an ANOVA?
ANOVASMAD! An analysis of variance to see if means are different! We are looking at means, not variance!

The way it works is simple: the program looks to see what the variation (variance) is within the groups, then works out how that variation would translate into variation (i.e. differences) between the groups, taking into account how many subjects there are in the groups. If the observed differences are a lot bigger than what you'd expect by chance, you have statistical significance. If, there are only two groups, variation between groups is just the difference between the means.

To use an ANOVA, rather than a t-test, you need 3 or more levels of one IV.
ANOVA Group Variance
Variance in the DV associated with each level of the IV. GOOD VARIANCE
ANOVA Error Variance
Variance in the DV associated with anything but the IV. or Within-groups variance (extraneous variables, sampling error).
F-statistic
F has a skewed distribution.
F will never have a negative value.
The farther F is from zero, the more significant it is.
When would you use a factorial ANOVA?
When you have 2+ IVs. Notation is number of IVs by number of levels, ex. 2x3.
Factorial ANOVA: main effects vs. interaction effects
Main effects: the effect of each IV on the DV independent of the other IV.
Interaction effects: the effects of one IV on the DV are dependent on the level of the other IV.

**Don't interpret main effects if there is an interaction effect because the results depend on the various conditions presented by the IVs.
Simple regression
Using a best fit line to approximate the relationship between one IV ( or predictor variable) and one DV (criterion variable). This would then be used to predict values for which you do not have actual data.

**You are essentially looking for slope, or the percentage of change in one varaible that is associated with the change in the other variable. Take any r, ex. r=.4. Get the variance by squaring it = .16, and then turn it into a percentage = 16%. Ie. 16% of the change in x is associated with the change in y.
Multiple regression
In multiple regression you are contrasting 2 or more predictor variables with 1 criterion variable. You are now considering not a 2 dimensiaonal graph but a 3+ dimensional graph.