PSYC 310 Final Exam

What are the 3 key relationships that are obtained in data from a two-way design?

- (1) will tell you if your first IV has a significant main effect on the DV - (2) will tell you if your second IV has a significant main effect on the DV - (3) will tell you if the interaction of the two IVs has a significant effect on the DV.

What is a One-Group Posttest-Only Design?

- A design in which a single group of participants receives the experimental condition and is then measured on the DV. - It's problem is that it has no comparison group.

What is a Posttest-Only w/ Nonequivalent Control Groups?

- A design in which one group receives the treatment condition and is compared to a group that did not receive the treatment condition. - It's problem is that it has Selection bias, Mortality differences, etc.

Define Systematic Sample

- A probability sample that is not randomly selected. - However, if the list has no structure to it (e.g., alphabetical), the results will be as good as random. These are practical because they require less work than a random sample. - (Example. You need 25% of the class for a survey, so you choose every 4th person on the student roster. (This is not random because numbers 1, 5, 9, etc had a 100% chance of being selected, and everyone else a 0%))

What are case studies?

- Case studies are a study of a situation which presents itself for investigation. - They often use a variety of sources such as interviews, documents, test results, and archival records, tries to explain unusual events/behavior in terms of established scientific principles. - it's great for developing and refining theories

Describe One-way Between-Subjects Anova

- DV is quantitative - interval level IV is between-subjects (3 or more different groups) - IV has 3 or more levels.

Describe One-way Within-Subjects Anova

- DV is quantitative - interval level IV is within-subjects (3 or more different measurements of the same people) - IV has 3 or more levels.

Describe repeated measure t-test

- DV is quantitative - interval level IV is within-subjects (comparing time periods) - IV has two and only two levels

What are factorial designs?

- Factorial designs are experiments which manipulate more than one IV (factor) at a time (two or more independent variables).

What is the difference between a regular latin square and a balanced latin square?

- In a regular latin square *everyone* goes from 1-2-3... always up by one, in the exact same order. - (example. so a subject may not start at the same number, but for everyone a 2 is always after a 1, and a 3 is always after a 2). - While in a balanced square with only a few people, every combo possible happens somewhere ("X comes after Y") - (example. Someone went from a 1 to a 2, someone went from a 1 to a 3, someone went from a 1 to a 4, someone went from a 5 to a 4... someone went from a 5 to a 1... Etc)

Describe independent measures t-test

- In it DV is quantitative - interval level IV is between-subjects (comparing groups) - IV has two and only two levels

Define Haphazard sampling

- Is when researcher uses hit or miss methods (in other words the researcher does not intend to employ a systematic approach to selecting a sample) - (eg. a TV crew interviews 10 people on the street) These are almost worthless scientifically.

What is a multiple condition within subject design?

- It is a design that has several levels (more commonly not just 2) of the IV, as well as a control condition.

What is a One Group Pretest-Posttest Design?

- It is a design that measures behavior of a single group before and after treatment. - The problem is that it has history, Maturation, Testing, Regression, Mortality, etc.

What is a 2 condition between subjects design?

- It's a design in which subjects are randomly allocated to only one group (either an Experimental or Control). - It can be modified for multiple groups/levels of IV.

Define Visual Analog Scale

- It's a straight line with the left end of the line representing least of something and the right end of the line representing the worst of something. - (For example, Patients are asked to mark on the line where they think their pain is.)

Define Complete Counterbalancing

- It's a way of controlling order within a group. - Each condition occurs an equal number of times in each rank-order position and follows each other condition an equal number of times.

What is a latin square?

- Latin Square is incomplete counterbalancing in which each condition is experienced by one subject in a systematically different order from other subjects. - Controls for order but not for sequence, and each condition appears only once in each row and column.

What is a likert scale or when is it used?

- Likert scale is used when question ask for a rating of the extent of agreement or disagreement with a statement. - Usually no more than 7.

What does Exhaustive mean?

- Means "an answer for every taste". - Basically all cases will fall into one of the alternatives, and the option "other" covers the outliers. - The other option can be bad though if a lot of people pick it

Why mix up questions in a questionnaire?

- Mixing them up will reduce respondent bias and require them to slow down and read each item. - Rather than just responding the same to a bunch of items written in one direction.

Define Close-ended questions

- Questions answered are limited to a set of choices. - They are easier to code and analyze but are too complex to reduce to a small set of alternatives and respondents may misinterpret or not agree with the given answer choices.

Define Open-ended questions

- Questions answered in your own words. - Yet, they are hard for researchers to code, and more tiring for respondents to answer.

Define Probability sampling

- Researcher knows the probability that any given individuals will appear in the sample. - Every member of the population has a "equal" chance of being selected, and this sample allows a researcher to use various statistical tests.

Define interaction effects

- The effects of combining the two factors together. - It is present when the relationship between one factor and the dependent scores depends on the level of the other factor.

Define Social Desirability Tendency (Beware of it)

- The tendency of people want to look good (or at least don't want to look bad)! - So every question has to be worded so that each alternative appears equally socially desirable.

What is Acquiescence (Beware of it)

- The tendency to agree with a statement on a questionnaire regardless of its Content. - (For example, People will agree with a statement and its opposite: "I like fish" ... and later ..."I don't like fish")

Define Numerical Scale

- Used when a person is asked to identify how much of something they are having or experiencing by choosing a number - (for example pain: from 0 (no pain) to 10 (the worst pain imaginable)).

Define Pain Faces Scale

- Uses 6 faces with different expressions ranging from happy because he/she has no pain, to sad because he/she has some or a lot of pain. - (ex. Patient is asked to choose the face that best describes how he/she is feeling. Used by people aged 3 years and older.)

Why do non-experimental research if it can't prove causality?

- You should do it because of the Hermeneutic Approach. - It asks what the behavior means rather than why it happened, and it sometimes captures the issue better than true experiments.

How many many observations are in the pretest-posttest design?

2

How many types of questions are there in a questionnaire / survey?

2

How many probability samples are there and what are they called?

2, and they are called Simple random, and stratified random sampling.

Define Verification key

A collection of items on a questionnaire designed to detect dishonest answers. - (Ex. Have you ever stolen anything, no matter how small?)

Define Convenience sampling

A non-random sample chosen for practical (mostly easy access) reasons

Define Purposive sampling

A non-random sample, chosen for some characteristic it possesses (eg. student athletes). - With this you have to be sure to select sample very carefully to avoid bias.

Define condition = Treatment

A particular group (grouped in terms of the way that the subjects experience the IV).

Define Sampling frame

A population as it is defined for the purposes of selecting participants for a study (eg. Students of Pepperdine = full-time, credit earning, students)

Define Population

A population is all of the people (or animals) in the group to whom the research applies

Define Multi-Stage Sampling

A sophisticated form of cluster sampling. Clusters are further broken down by taking samples from within each cluster. - (Example. Gallup poll - They randomly select several zip codes, then randomly select streets within those zip codes, then randomly select addresses on those streets, then they give the poll at those addresses. Not as accurate as random sampling - but with a sample size of 1000, you can estimate within +/- 4%)

Define Elements

A term to describe each individual that falls within the sampling frame. - (a # of elements from the sampling frame is the number of people)

Define Simple random sample

A type of probability sample in which each member of the population has the same chance to be chosen, independently of any other member

Define Main effects

All the information about the effect of each factor alone.

Define Sequence Effects

Conditions interfere with each other.

Define Counterbalancing

Counterbalancing is when subjects experience the conditions in a variety of orders.

What are some limitations of archival research?

Data was collected for some other purpose, making ruling out alternative hypotheses difficult, no real real possibility of informed consent, and the accuracy of the previous data is hard to determine.

What are factors?

Factors are another word for IVs. - There is always at least one factor in an experiment, therefore a Single-Factor Design means only one IV.

What are the two tests of Single-Factor Designs (2 levels)?

Independent measure t-test and Repeated measures t-test

Define Block Randomization

Is when the order of conditions is randomized, but each subject is presented with each condition before any condition is repeated.

How is participant observer research ethically controversial?

It can invade participants privacy, and insider view bias may result.

Facts about non-experimental research?

It is frequently qualitative, and often called correlational research.

What is a 2 condition within subject design?

It is the Simplest true-experimental design, in it subjects experience both experimental and control conditions, and the conditions are conditions counterbalanced across subjects AB or BA.

What are the disadvantages of a case study?

It is time consuming, can result in potentially inaccurate conclusions, its generalizability is questionable, and does not tell us about cause and effect.

Define Naturalistic Observation

It is unobtrusive or nonreactive research that is sometimes referred to as fieldwork or field observation.

Why not mix open ended and closed questions within a survey?

It makes it easier to analyze, yet provides insight into answers by allowing respondents to expand on the answers to closed-ended questions.

Why is random assignment such a big deal?

It's a big still because In order to logically conclude that our manipulation made a difference, we need to be able to state that the groups were the same to begin with.

What are the advantages of a case study?

It's important in studying rare characteristics or events, and can provide a counter instance of an otherwise well- supported theory.

What is participant-observer research?

It's research in which an observer joins a group in order to record behavior. Most effective in small and remote groups

What is a balanced latin square?

It's special design of a Latin square that controls the sequence effect of the immediately preceding condition.

Define Reverse Counterbalancing

It's when conditions are presented first in one order, then in the reverse.

Where does laboratory observation occur?

Laboratory observation happens in the lab instead of the field.

What does Mutually exclusive mean?

Means no case can belong to more than one alternative at a time (ie: male or female)

Define Non Reactive

Means subjects should be unaware that they're being studied.

Define Latent Content

Measures the appearance of themes, as interpreted by the researcher.

Define Manifest Content

Measures the frequency of some word, image, phrase or action.

Name other Non Random Samples

Multi-stage Sampling

What are the 4 types of non-experimental research?

Observational (which means observe ongoing behavior), Archival (which means examine existing data), Case Study (which means study existing situation), and Survey/Interview (which means assess traits using questionnaire).

What are the two Single-Factor Designs (3 or more levels)?

One-way Between-Subjects Anova and One-way Within-Subjects Anova

Name the 3 measures of naturalistic observation

Physical trace measures, erosion measures, and accretion measures.

Define Branching Items / questions

Questions that enable the participant to flow through the survey in an individualized way, depending on his/her answers to certain questions

When controlling order & sequence effects you should?

Randomize the order of conditions, block randomization, and reverse counterbalancing.

What is archival research?

Research attempting to answer research questions from archival data, which is factual information in preexisting records or archives.

What is non experimental research?

Research in which the experimenter does not manipulate an IV, does not assign subjects to groups, and does not have control over the conditions of the study.

Define Response Rate

Response rate is the percentage of individuals who return a completed survey. - (example: Most researchers require at least a 50% return rate before they consider a survey to be truly "representative" of a population.)

What happens in a true experiment?

Scientists have complete control over all aspects of the experiment (who, what, when, where, and how), subjects are randomly assigned to conditions, and there is a control group or a control condition.

What are the non random probability samples?

Systematic Sample, cluster sample, and multi-staging sample.

Define Content Analysis

The evaluation of a hypothesis using publicly available pictures and language.

When studying factors (independent variables) at once we get?

The main effects of each factor alone, and the interaction effect of the two factors together.

Define Order Effects

The order position of a condition changes a subject's performance.

How many types of samples are there and what are they?

There are 4, and they are called Haphazard sampling, Purposive sampling, Convenience sampling, and Probability sampling.

Define levels

They are the different values of a factor.

What type of question does a 2 condition within subject design answer?

This design answers a yes/no question (example: Did the treatment have an effect?).

True or False: it is good to have two coders for all observational recordings for reliability

True

True or false: each factor can contain any number of levels

True

Define Physical trace measures

Unobtrusive measures of physical evidence.

What do you do if you don't have enough subjects to do complete counterbalancing?

Use a latin square.

Define Stratified Random Sample

When two or more subsamples are represented according to some predetermined proportion - (usually in the same proportion as they exist in the population).

What happens in within-subjects with true experimental designs?

With within-subject designs, all of my subjects experience the same conditions (they're all in the same group), and the order of the conditions should be counterbalanced across subjects.

Can you create a Coding guide?

Yes

Is putting your questions in the right order important when designing a questionnaire?

Yes

Will you be able to pick your statistics?

Yes

Does sequence matter when designing a questionnaire?

Yes, particularly when asking branching questions.

When designing a questionnaire how should you Mix It Up?

You can mix it up by (1) using statements that are both positive and negative for the trait, (2) using statements stated favorably and unfavorably, (3) or reverse the direction of the options.

How many main effects will you have for an experiment with 3 IVs?

You will have three, the main effects of A, B, and C.

When performing naturalistic observation

carefully keep records, have a checklist, use a variety of measures, and remember privacy is important.

Define Pre-existing data

data that was collected for some other purpose. Simply examines or selects the data for analysis.

Define Accretion measures

evidence that is left behind

When taking field notes you should be?

systematic, selective, use recording devices, and don't forget that privacy is important.

Define Unobtrusive

watch behavior, but don't try to change it.

Define Erosion measures

wear and tear