Term
|
Definition
| Systematic Empiricism, Publicly Verifable Knowledge, Investigates Testable Ideas, Predictions and Theories must be Falsifiable, and Tentative Progress. |
|
|
Term
|
Definition
| Knowledge gained through controlled organized observation rather than pure thought, common sense, and authority. |
|
|
Term
| Publicly Verifiable Knowledge |
|
Definition
| Recipes for observations made so they can be replicated. Ideas are sumbitted for public review (published in books or journals, peer review process) |
|
|
Term
| Investigates Testable Ideas |
|
Definition
| That is problems for which observation can clearly decide an answer, such as is smiling contagious. But not like, what is the meaning of life (those questions cannot be answered) |
|
|
Term
| Theories and Predictions must be Falsifiable |
|
Definition
| In a scientific theory predictions must be able to be right or wrong. Must be falsifiable or refutable. (Non-Falsifiable: Irrefutbale hypothesis which is a hallmark property of a pseudoscience) |
|
|
Term
|
Definition
| More powerful theories make more specific predictions. The more specific they are more easily falsified. |
|
|
Term
| Sciene is Tentative and Progressive |
|
Definition
| Science is a series of conjectures and refutations. You propose then their falsified and you start the whole process over again. Its a gradual progression. |
|
|
Term
| Science vs. Pseudoscience |
|
Definition
| Science verifies conclusions through controlled observations and replication, predictions must be falsifiable and theories are tentative. |
|
|
Term
|
Definition
| Pseudoscience verifies conlcusions through anecdotal evidence, predictions are irrefutable, and refusion to revise in light of evidence. |
|
|
Term
| What makes a good research question? |
|
Definition
| Needs to be empircally testable, and should be newsworthy. |
|
|
Term
|
Definition
| Title Page, Abstract, Introduction, Method, Results, Discussion, Reference, Tables and Figures (one per page) |
|
|
Term
|
Definition
| Interrelated concepts that explain a body of data |
|
|
Term
|
Definition
| Predicted answers to question. Relationship between conceptual variables. |
|
|
Term
|
Definition
| quality type characteristic like Gender, a nominal or category scale (category) |
|
|
Term
|
Definition
| How much of something (like Emotional control), ratio or interval scale |
|
|
Term
|
Definition
| scale has a 0 and it means the absence of the property (like measuring emotional control by brainwaves) |
|
|
Term
|
Definition
| scale doesnt have a 0 and if it does it doesn't indicate the absence of the property (rating scale 1-10) |
|
|
Term
| Two kinds of questions to ask in a questionairre: |
|
Definition
Free-format=easy to think up but hard to quantify (Ex How would you react if your bestfrined told you she had a "fling" with your boyfriend)
Fixed-Format= hard to create but easy to analyze; uses the likert scale |
|
|
Term
| Problems with Self-reports: |
|
Definition
| Reactivity- subjects know that they are being measured |
|
|
Term
|
Definition
| Unsystematic and Systematic error |
|
|
Term
|
Definition
Is a threat to reliability, which is undermined by random error. Random error such as coding errors, participants inattention to and misperception of questions; like noise.
(realiability is typically a concern for surveys but also coding) |
|
|
Term
|
Definition
| Is a threat to construct validity and is influenced by other conceptual variables like slef-esteem,mood, self-promotion. |
|
|
Term
|
Definition
| The extent to which a measured variable actually measures the conceptual variable that it is designed to assess. |
|
|
Term
|
Definition
| appears to measure variable |
|
|
Term
|
Definition
| measures all aspects of the concept |
|
|
Term
|
Definition
| Does a measure correlate with other measures of the concept (Two surveys that measure iQ should correlate across subjects) |
|
|
Term
|
Definition
| A measure should not correlate with "different" measures (a social competence measure should not correlate with a social dominance) |
|
|
Term
| Two parts to design validity: |
|
Definition
|
|
Term
|
Definition
| Does the research design really address the research question |
|
|
Term
|
Definition
| do results generalize beyond research setting |
|
|
Term
| The difference between Correlational and Experimental designs |
|
Definition
| Correlational measures variables to see if they relate, (correlate) and uses a regression analysis. Experimental manipulates circumstances and measures variables to see if they relate and uses Analysis of variance (ANOVA) |
|
|
Term
| Correlational and Experimental deisgn (cont'd) |
|
Definition
| Correlational has a predictor and a criterion whereas a Experimental design has an IV & DV. |
|
|
Term
| How do you increase internal validity? |
|
Definition
| by eliminating confounding variables and confounds are alternative explanations for why the DV changes and with them the experimental design doesn't really address the research question. |
|
|
Term
| How do you increase statistical validity? |
|
Definition
| by eliminating extraneous variables (noise) |
|
|
Term
|
Definition
| Conclusions could be incorrect because of Type I or II error |
|
|
Term
| To eliminate confounding variables: |
|
Definition
| Equate groups on confounding property, make confounding property an IV. |
|
|
Term
| To eliminate extraneous variables: |
|
Definition
| Eliminate individual differences by using limited populations, and eliminate experimental errors by standardizing the conditions (make sure the procedures are done exactly the same way for each subject) |
|
|
Term
|
Definition
| By doing a pilto study it helps to check for experimental error, confounds, and strength of manipulation. |
|
|
Term
|
Definition
| Subject bias in which the research hypothesis is apparent to the subject and the subject reacts in a certain way. To avoid this make the subject "blind" to the condition. Experimenter bias the same thing applies here but this time its the experiment who treats subjects a certain way. |
|
|
Term
| What reduces external validity? |
|
Definition
| Control; if you control results may not generalize to other settings and subjects. |
|
|
Term
|
Definition
| tend to have more external validity and less internal validity |
|
|
Term
|
Definition
| tend to have more internal validity and less external validity |
|
|
Term
|
Definition
| Increases external validity because any one single research hypothesis will always be limited in terms of what it can show. Advances in sciene occur through the accumulation of knowledge that comes from different tests of the same theory which is where replication lies. |
|
|
Term
| Six APA ethical guidelines (first three): |
|
Definition
| 1. Perform research with concern for dignity and welfare of participants. 2. Obtain informed consent from participants (person is not a subject but a participant, needs to be in writing). 3. Freedom from coercion (they must be told that they can "withdraw at any time"). |
|
|
Term
| cont'd. Six APA guidelines (last three): |
|
Definition
| 4. Use deception only when necessary. 5. Confidentiality/Privacy (participants should not be associated with their name). 6. Debriefing. |
|
|
Term
| When is it ok to use deception: |
|
Definition
| It is ok to use when it is about the process being investigated but its not when you provide false information about their performance. |
|
|
Term
| Confidentiality vs. Anonymity |
|
Definition
| Anonymity you can't figure out who the subject was whereas confidentiality you can. |
|
|
Term
|
Definition
| participants must have an opportunity to obtain info about then nature and results of the research, debriefing should include a question and answer period, and all deceptions must be explained. |
|
|
Term
Research Design: Non-Experimental
(kinds) |
|
Definition
| Naturalistic observation, participant observation, and surveys. Which can also be broken down into observational and survey desgins. |
|
|
Term
| Naturalistic Observation: |
|
Definition
| Observe behavior in its naturally occurring environment with not attempt to control or manipulate. These studies have ecological validity. Observations should be unobtrusive. |
|
|
Term
|
Definition
| Is the research setting like real life |
|
|
Term
| Problems with Naturalistic observations |
|
Definition
| No control hence hard to darw conclusions about causes... |
|
|
Term
|
Definition
Type:Ethongrapy
Observe behavior in its naturally occurring environment.
Observer becomes immersed into a behavioral or social system.
The problem with it is reactivity and jnvasion of privacy. |
|
|
Term
|
Definition
| Identify research question, define conceptual variables to be measured, develop questions to measure variables. |
|
|
Term
| Kinds of questions for questionnaire: |
|
Definition
Demographic= social properties like age, sex, income.
Open-ended= answer questions in own words.
Close-ended= limits answers to certain alternatives (Ex. where do you drink? dorm, apartment, bar.)
Partially opend-ended= its close-ended but you add "other", providing a little more freedom. |
|
|
Term
| Format for Close-ended questions: |
|
Definition
| Multiple choice, true/false, and rating (ex. likert scale) |
|
|
Term
| What makes a good question? |
|
Definition
| precise address a single issue, understandable language, and unbaised. |
|
|
Term
|
Definition
| threatens reliability consistent results on repeated administration, and validity measuring what you intended. |
|
|
Term
| How do you order questions? |
|
Definition
| Put demographic questions last, put related questions together, and intermix objectionable and non-objectionable questions |
|
|
Term
|
Definition
| samples should be representative of population and each person has equal probability of being chosen. |
|
|
Term
|
Definition
| When the sampling frame is not accurate or doesn't exist (eg. homeless) |
|
|
Term
| Nonprobability sampling techinques: |
|
Definition
Snowball sampling- intial subjects led experimenter to other subjects (friends).
Convenience sampling- use whatever subjects are available. |
|
|
Term
| The difference between non-experimental and experimental designs: |
|
Definition
| With non-experimental which are made up of observe and self-report; because there is no manipulation there is cause for possible confounds. Experimental designs manipulate conditions. |
|
|
Term
| The experimental design... |
|
Definition
| manipulates circumstances to get temporal ordering right and control confounding variables |
|
|
Term
| Two vs. multiple condition experiment |
|
Definition
Two condition= contains experimental and control group, t-test.
Multiple condition= has more than one control and experiment group, ANOVA. |
|
|
Term
| Between vs. Within group differences |
|
Definition
Between= variability due to differences between conditions, systematic variability (=IV or confounds)
Within= variability due to diff. within the conditions, unsystematic variability (=Random nosie) |
|
|
Term
| Repeated measures (aka within-subject design) |
|
Definition
| Use same subjects in the different conditions |
|
|
Term
| Advantages & Disadvantages to Repeated measure |
|
Definition
Advantage- makes statistical test more powerful by controlling for individual differences and decreases within group variability.
Disadvantages- carry-over effects, performance on later conditions is affected by previous exposure (practice, fatigue) |
|
|
Term
| Solution to carry-over effects |
|
Definition
| Counterbalancing- arranging the order of conditions in a different way to equate them |
|
|
Term
| Quasi-experimental vs. Experimental |
|
Definition
Quasi- subjects are assigned to levels by vrtue of their pre-existing characteristics (which doesn't eliminate possible confounds)
Experimental- subjects are randomly assigned to the different levels |
|
|
Term
| One-way vs. Factorial design |
|
Definition
| Factorial there is more than one IV whereas one-way has only one IV. |
|
|
