Term
| Why study research methods? L1 |
|
Definition
- to understand the source of much CHYS knowledge
-survive upper year courses!
-to have an edge in the workplaces |
|
|
Term
| Authority (Ways of knowing)L1 |
|
Definition
-scientific method is the only way of knowing
- simplistic, but quality of information related to quality of source.
- every research study uses this method. |
|
|
Term
| Rational Method (Ways of knowing)L1 |
|
Definition
- Relies on past knowledge plus logic
- used when direct observations dangerous, impractical, unethical
-every research study uses the rational method. |
|
|
Term
| Empirical Method (Ways of knowing)L1 |
|
Definition
- Sense experience (taste, touch, smell, sight, hearing)is the only basis for knowing
-can range from simple observation to more complicated. |
|
|
Term
|
Definition
1) Begin with an observation
2) Form a hypothesis
3) Make a prediction
4) Conduct observations
5)Based on observations, form a new hypothesis. |
|
|
Term
| How do you go about changing topics into numbers? L2 |
|
Definition
| -Pretest results, stem and leaf, box plot, histogram |
|
|
Term
|
Definition
| Did you measure what you said you would measure? |
|
|
Term
|
Definition
|
|
Term
| Nominal (Levels of Measurement) L2 |
|
Definition
Gives a name to data points. (identity)
-Measurement does not assign people to groups but an identity. i.e., male/female, attachment theory. |
|
|
Term
| Ordinal (Levels of Measurement) L2 |
|
Definition
| Orders data points, presented in order of complexity (rank order). Does not give degree/scale. |
|
|
Term
| Interval (Levels of Measurement) L2 |
|
Definition
| Specifies the distance between increments. |
|
|
Term
| Ratio (Levels of Measurement) L2 |
|
Definition
| like interval, but has a true zero (absence). |
|
|
Term
|
Definition
| Nominal, Ordinal, Interval, Ratio (Levels of Measurement) |
|
|
Term
| A typical self-report Scale. L2 |
|
Definition
-multiple items targeting the same construct
-participant selects how true each statement is |
|
|
Term
|
Definition
- a statement a participant is asked evaluate construct.
-Referenced against subjective criteria
- 5 gradations of endorsement
- multiple items are summed/averaged and treated as an interval scale |
|
|
Term
|
Definition
-Research version of concept/idea
-idea and how it is measured. |
|
|
Term
| General Research Ethics L3 |
|
Definition
1) Respect for human dignity
2)Free and informed consent
3)Respect for vulnerable persons
4) Privacy and confidentiality
5)justice and inclusiveness
6) Balancing harms and benefits |
|
|
Term
| 1st principle of modern research ethics: L3 |
|
Definition
| -Protect all the interest of participants: bodily, psychologically, and cultural |
|
|
Term
| Lack of protection towards participants are demonstrated by what group/study L3 |
|
Definition
| Nazis experimentation (concentration and science camps) |
|
|
Term
| Lack of free and informed consent demonstrated by what groups/study (2) L3 |
|
Definition
| Tuskegee Syphilis Study and Thalidomide Incident |
|
|
Term
| Two reasons why free and informed consent can be difficult. L3 |
|
Definition
1) Not just simple as agreeing to terminate
2) Complicated for children, mentally incompetent, those institutionalized. (vulnerable) |
|
|
Term
| Give one study that demonstrates the special considerations of vulnerable persons being ignored. L3 |
|
Definition
|
|
Term
| Confidentiality (Related to Research Ethics) L3 |
|
Definition
| -Researchers must ensure there is no link between personal info and identity. |
|
|
Term
|
Definition
| -No one group unfairly burdened with the harms of research |
|
|
Term
| How to balance harms and benefits (research study) L3 |
|
Definition
-foreseeable harms should not outweigh benefits
-minimizing harm: participants should be subjected to minimal harm
- any harm must be essential to scientific and societal good. |
|
|
Term
| How to achieve informed consent in research study L3 |
|
Definition
1) Assessment of participants understanding and appreciate of implications
2) assurance that the participant has the freedom to chose without repercussion. |
|
|
Term
|
Definition
| -truly random assignment i.e. coin toss. |
|
|
Term
| Systematic Random Sample L4 |
|
Definition
| -numbers assigned, rules for selecting participants |
|
|
Term
| Stratified Random Sample L4 |
|
Definition
- defined number per group
- downfall: not able to generalize to population |
|
|
Term
| Proportionate Stratified (Random) Sample L4 |
|
Definition
| - sample according to size of each group in the population |
|
|
Term
| Cluster (Random) Sample L4 |
|
Definition
target population (i.e. kids in classrooms), each cluster different
Random Sample of individuals within cluster. |
|
|
Term
| Descriptive Research Study L4 |
|
Definition
| rare, usually done on things that are little known. Tables are common. |
|
|
Term
| How to identify a experimental or quasi-experimental study L4 |
|
Definition
1) Random assignment of individual to conditions (method sections)
2) Conditions may be called: groups, treatments, interventions, scenarios etc.
3) Conditions must be created by the research if not then Quasi.
4) Rules out alternate explanations
5) High internal validity, sometimes low external validity. |
|
|
Term
| How to identify a correlational/non-experimental study. L4 |
|
Definition
-Examines the relationship between 2 variables
- Look for: "Correlated with" "Associated with" etc.
-co-relation matrix = co-relational study
- in a correlational study look for statistics.
- cannot draw conclusions about individuals only constructs |
|
|
Term
|
Definition
| Are the findings to be trusted? |
|
|
Term
|
Definition
| Are the findings to be trusted due to how closely the study reflects reality in the population |
|
|
Term
|
Definition
| Are the findings to be trusted based on experimental procedures and design. |
|
|
Term
|
Definition
-The variable other things influence
- Construct researchers are trying to understand
-measured through frequency, intensity, or duration |
|
|
Term
|
Definition
- First manipulation of DV.
-Thing you want to effect DV
"Predictor Variable" |
|
|
Term
|
Definition
- Measurement and 2nd measurement of DV
- How do you know that you are measuring what you are claiming to? |
|
|
Term
|
Definition
| Anything that influences scores on DV that isn't IV. |
|
|
Term
| Limits of Control for Extraneous Variables L5 |
|
Definition
1) Random Assignment
2)Holding Variables constant
3) Control Groups |
|
|
Term
| Basic Structural Differences of Between-Subject Designs. L6 |
|
Definition
1) Different Participants for different conditions
2)Every person participates in only one characteristic |
|
|
Term
| What is a major problem with Between-Subject Designs. L6 |
|
Definition
-Individual Differences (i.e starting mood)
-makes it hard to get two even groups. |
|
|
Term
| Why use Between-Subject Designs? L6 |
|
Definition
| -Score independence: Any changes in one group did not happen in every group. |
|
|
Term
|
Definition
| -Every participant participates in every condition. |
|
|
Term
| What is a major problem with Within-Subject Design? L6 |
|
Definition
| - Exposure to one condition is going to effect exposure to 2nd condition |
|
|
Term
| How to identify Between-Subject Design L6 |
|
Definition
-conditions i.e. mood variant
-creation of condition
-research control over the effects
-manipulation check |
|
|
Term
| Why are Within-Subject Designs generally used? L7 |
|
Definition
Generally used when problems of design can be avoided, then better then between-subjects
-Little difference in mood, controls variance
|
|
|
Term
| How to identify experimental Within-Subject design. L7 |
|
Definition
-True independent variable: Researcher creates categories
-Research has control over effects of IV
- may need manipulation check
Quasi IV:
-Categories found not created
-No control over who is in what category
-susceptible to multiple confounds |
|
|
Term
|
Definition
| Measuring participants on 1st level effects 2nd level etc. |
|
|
Term
|
Definition
| When event outside experiment affects DV. |
|
|
Term
| Solutions for Carryover effect in Within-Subject Design L7 (3) |
|
Definition
1)Don't use WS. Design
2) Use WS. Design but use this weakness as a strength
3) Turn WS. factor into a BS. factor and do matching procedure. |
|
|
Term
| When is Matched-Subject Design Used? L7 |
|
Definition
| -Used when cost of WS. design are too high. |
|
|
Term
| What is an advantage of Matched-Subject Design L7? |
|
Definition
| -Gives the control/sensitivity of a WS. design without downfalls. |
|
|
Term
| What are the disadvantages of Matched-Subject Design? (3) L7 |
|
Definition
1) Tedious and wasteful
2) Must identify a matching variable
3) Interpretation problems due to biased sampling. |
|
|
Term
| What is counterbalancing and what is it a solution for? L7 |
|
Definition
-Giving each participant a different sequence of the different IV levels.
-Solution: Sensitivity due to carryover effects. |
|
|
Term
| When is counterbalancing effective? L7 |
|
Definition
- Effective if there only a few levels (up to four)
- more levels prevents all combination of levels from being used. |
|
|
Term
| What is the use of a Latin-Square? L7 |
|
Definition
|
|
Term
| Demonstrate an example of a basic Latin square, using ABCDE L7 |
|
Definition
Correct Answer:
-ABCDE
-BCDEA
-CDEAB
-DEABC
-EABCD |
|
|
Term
| Give the formula, and demonstrate an example of balanced Latin-square design using 12345. L7 |
|
Definition
Formula: 12n 3n-1
Correct Answer:
12534
23145
34251
45312
51423 |
|
|
Term
|
Definition
| 1 independent variable that divides participants in group |
|
|
Term
|
Definition
| More then 1 IV variable and the independent are crossed with each other |
|
|
Term
| Explain what the numbers in factorial design represent (i.e. 2x2) L8 |
|
Definition
| - Number of Numbers refer to Number of IV. Value of number represents the number of levels for that IV. |
|
|
Term
| Where are the main effects of differences among the levels of one IV found? L8. |
|
Definition
|
|
Term
| How do you report the statistic used to compare means with what symbol. L8 |
|
Definition
|
|
Term
| How do you report the effect size and what is it? L8 |
|
Definition
d = or n2
-how big/important the result is. |
|
|
Term
| How do you report the probability that a result as big as yours would be found by chance? L8 |
|
Definition
|
|
Term
| What P value is considered significant? L8 |
|
Definition
|
|
Term
| Interaction (related to factorial design) L8 |
|
Definition
If one effects changes depending on other IV.
- an interaction happens when there is a different |
|
|
Term
| Assumptions/ requirements of factorial designs. L8 |
|
Definition
1)Independence: Each participant only in 1 condition on each IV.
2) Normality: Data within each cell of the design must be normally distributed
3)Homogeneity of variance: all cell must have similar variances. |
|
|
Term
| What does a Factorial ANOVA answer? L9 |
|
Definition
| Was the independent variable significant? What were the effects? |
|
|
Term
| Give the different categories of n2? L9 |
|
Definition
small .01
medium .05
large .10 |
|
|
Term
|
Definition
Participants randomized to IV. Conclusion: Unobserved variables controlled
-fewest threats to internal validity |
|
|
Term
|
Definition
-Pre-existing groups compared
Conclusions: Independent Variable might have caused Dependent Variable
-Multiple threats to Internal Validity |
|
|
Term
|
Definition
| Groups formed carefully. Conclude: IV likely caused DV. |
|
|
Term
| Why use Differential Research? L9 |
|
Definition
-When participants cannot randomly assigned
-difficult/low-frequency population
- when it is unethical to assign groups |
|
|
Term
| 1 disadvantage and 1 advantage of use differential research? L9 |
|
Definition
-Advantage: Can do studies that otherwise couldn't be done
-Disadvantage: Can't trust findings about group differences due to multiple internal validity problems. |
|
|
Term
| What is a Post Test only non-equivalent control group design (PONCGD)? L9 |
|
Definition
| -2 groups, measured after 1 group receives treatment etc. |
|
|
Term
| How can a PONCGD be improved? L9 |
|
Definition
- 1) Randomizing prior to assignment
-2_ Matching on Control Variable. |
|
|
Term
| Pretest- Postest Non-equivalent Control Group Design (PPNCGD) L9 |
|
Definition
| -2 groups, measured before and after 1 group receives treatment |
|
|
Term
| How can an PPNCGD be improved? L9 |
|
Definition
| - 1) Randomizing prior to assignment matching prior to assessment. |
|
|
Term
|
Definition
-Using time & known events to sort out causality
- 1 group measured for time, then condition is introduced, then measurement continues
- If the changes occurs with the even then the event has caused the change |
|
|
Term
| Problems with a Time-Series Design L9 |
|
Definition
1) Change takes time to occur
2) Change is not consistent
3) Other events occurring
4) How much change is enough? |
|
|
Term
| What can be used if there are 2 variables and the concern is their relationship and what does this figure depict? L10 |
|
Definition
| Scatterplot, when interpreted portrays the usually imperfect relationship |
|
|
Term
| Three things important to look for in scatterplots. L10 |
|
Definition
Direction (Positive or Negative)
-Magnitude/Strength: Depending on Scatter
-Shape/Form-Linear most important. |
|
|
Term
| Pearson, requires what. L10 |
|
Definition
- Linearly related data
-Influenced by outliers/extreme scores
-ranges from -1 to 1
-Directional and Non-directional predictions
- Described with the coefficient of determination (r2) |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
Kick -> <- OW
- both cause each other. |
|
|
Term
| Spurious Relationship L10 |
|
Definition
|
|
Term
| Extraneous Variable Relationship L10 |
|
Definition
|
|
Term
| Describe CLPCD (Cross-lagged Panel Correlational Design) L10 |
|
Definition
- 1 method of ruling out non-causal relationships
-need longitudal data with repeated measurements of 2 or more points. |
|
|
Term
| Three things needed to understand a set of scores L11 |
|
Definition
1) Distribution/Shape
2)Central Tendency
3) Dispersion/Spread |
|
|
Term
| Frequency Distribution Graph L11 |
|
Definition
-orders data highest to lowest
= x gives value
- gives you n at each value and p% for each
-says nothing about scores as whole |
|
|
Term
| Stem and leaf display L11 |
|
Definition
- Good technique to see every datapoint and shape info
-preserves info on proportions
-can have spilt stems if enough data, to show more precise shape info. |
|
|
Term
|
Definition
-Frequency Graph turned on side
- n in each x category
- if you have interval/ratio data bars will touched
- if data is nomina/ordinal, bar graph used instead |
|
|
Term
| Types of calculation for Central Tendency L11 |
|
Definition
| Mode, Median, Mode, Range, Variance, SD |
|
|
Term
| Strengths and Weaknesses of Mode L11 |
|
Definition
Strengths
1)Quick estimate of central tendency
2)Highest probability of being right
Weaknesses
1) Gives no info about distribution
2) Calculations depend on the level of measurement |
|
|
Term
|
Definition
| Calculation: Most common score. |
|
|
Term
|
Definition
Calculation: Need to find median location
-middle score |
|
|
Term
| Strengths and Weaknesses of Median L11 |
|
Definition
Strengths:
-1) Miminizes average distances between itself and other raw scores
2)not sensitive to extreme scores, but still sensitive to overall shape of distribution
Weaknesses:
1)May not correspond to an actual score
2) Need special formula if multiple medians |
|
|
Term
|
Definition
| Calculation: Sum of all scores divided by number |
|
|
Term
| Strenghts and Weaknesses of Mean L11 |
|
Definition
Strengths:
1) Value of every score used in finding the mean.
2) Least Biased measures of central tendencies
3)Greater weight to extreme scores
Weaknesses:
1) Restricted to interval or ratio data
2) Heavy influenced by extreme scores and the overall shape of distribution |
|
|
Term
| Variance Strengths and Weakness L11 |
|
Definition
Strengths
1) Gives an estimate of total variability in sample
2) The value of every score is use to calculate
3) Unbiased
Weaknesses
1) Not easy to interpret
2) Not a quick, intuitive calculation |
|
|
Term
| Standard Deviation (SD) L11 |
|
Definition
Strengths
- More intuitive than variance
- contrasts samples measure on same scale
-has interesting characteristics
Weaknessnes
-depends on the assumptions about shape
-not unbiased |
|
|
Term
|
Definition
"extension of co-relation"
-Regression helps you predict the relation to the DV, for someone not in the original study. |
|
|
Term
| Pearson describes what: L12 |
|
Definition
| - Strength and the direction of the relationship. |
|
|
Term
| How is the Coefficient of Determination displayed? L12 |
|
Definition
|
|
Term
| What is equation for regression and what does each figure represent? L12 |
|
Definition
Y = bX + A
- Y can also be Y1 or ^
Y
- Y is predicted DV for participant.
- The Y score is predicted by corresponding X score, multiplied by some number and added to another.
- b is the slope, tells how much the participants score changes when X changes by 1 unit.
-a is the intercept, and represents the individuals y score when x is zero, number may not always make sense. |
|
|
Term
| Complex equation for regression L12 |
|
Definition
| Y = b1 X1 + b2 X2....bk Xk + a, R2 |
|
|
Term
| Multiple coefficient of determination L12 |
|
Definition
|
|
Term
| What is a Standard Error? L12 |
|
Definition
| -The mean deviation you would expect if you ran the study an infinite number of times with samples of the same size. |
|
|
Term
| How do you calculate a standard error? L12 |
|
Definition
1)Need a Standard Deviation
2) Divide the SD by the square-root of N to get SE.
3) Multiply the SE by 1.60 by 1.96
4) Add the result of step 3 to the mean, set aside
5) Subtract the result of 3 to the mean set aside.
6) Graph |
|
|
Term
| Important things to remember when calculating Standard Error? L12 |
|
Definition
- SE works with multiple means but measures must be identical
- need to have standard deviations of all groups
- heterogeneous variances may confuse interactions
- small numbers with large SE.
- Using SE. to graph results, not statistical test. |
|
|
Term
|
Definition
-Every statistic has an effect size associated with it.
- Involves two means, it is nothing more than the mean difference divided by the standard deviation
- if it involves correlation, it is the correlation |
|
|
Term
| Effect Size Guidelines L12 |
|
Definition
- Small: d 0.20, 0.50, 0.80
r/f/n/w = 0.10 % distribution overlap: 85
- medium: d = 0.50, r/f/n/w 0.30, % distribution overlap 67
-large 0.80, r/f/n/w 0.50, % distribution overlap 53. |
|
|
Term
| Cautions on Effect size L12 |
|
Definition
-d doesn't work with multiple means
- needs to have SD of both groups
-must have homogeneous variances
-N doesn't matter but small N will make means unstable.
- |
|
|
Term
| When you “know” that you do not want to eat fried worms, even when everyone around you says that they taste great, your decision is based on |
|
Definition
|
|
Term
| A restaurant chef tried replacing rice with pasta in one of her recipes to see what would happen. Which method of acquiring knowledge is she using? |
|
Definition
|
|
Term
| A patient who demands a second opinion before agreeing to surgery, is double-checking information obtained by the |
|
Definition
|
|
Term
| Adolescents occasionally develop eating disorders. You wonder what is the best way to treat eating disorders. This is an example of getting research ideas from |
|
Definition
|
|
Term
| Which item is not provided in a detailed record for item in PsycInfo? |
|
Definition
| details of research methodology |
|
|
Term
| The method section of a research article typically |
|
Definition
| provides the details of how the research was conducted. |
|
|
Term
| It is possible for psychologists to study variables such as hunger, motivation, and self-esteem because these variables can be measured by |
|
Definition
|
|
Term
| A(n) ________ measure produces similar results when the same individuals are measured under identical conditions. |
|
Definition
|
|
Term
| An EEG, heart rate, and brain activity are all examples of ________ measures. |
|
Definition
|
|
Term
| Under which circumstance is informed consent not necessary? |
|
Definition
| a public opinion research where participants complete anonymous questionnaires |
|
|
Term
| Deception is in direct contradiction with which element of the ethical guidelines? |
|
Definition
|
|
Term
| Which form of deception is not permitted |
|
Definition
| not telling participants about a possible risk in the study |
|
|
Term
| The group of individuals from which researchers actually select participants for research studies is called |
|
Definition
|
|
Term
| Although a research question usually concerns the _________ the actual research participants are selected from the _________. |
|
Definition
| target population, accessible population |
|
|
Term
| A researcher asks students to sign up for a research study and then selects the first 15 males and 15 females who sign up. The researcher is using |
|
Definition
|
|
Term
| research study begins by separating a sample of 40-year-old men into two groups: a high income group and a low income group. A score measuring alcohol use is then obtained for each man. The study intends to determine whether there is a relationship between income and alcohol use. This study is an example of the ____________ research strategy. |
|
Definition
|
|
Term
| If a researcher finds that room color influences mood of the adults in a sample, then the researcher would like to conclude that room color affects mood of |
|
Definition
|
|
Term
| In a study with a group of individuals being tested in a series of treatment conditions that extends over a relatively long time, it is possible that there will be systematic changes in the participants’ physiology or psychology that occur during the time of the study. If these changes influence the participants’ scores, causing scores at the end of the study to be different from scores at the beginning, the effect is called |
|
Definition
|
|
Term
| If a study finds that there is a strong relationship between two variables, this means that |
|
Definition
| you cannot draw a conclusion about the causal relationship between the variables. |
|
|
Term
| A researcher decides to use only male participants in an experiment comparing two treatment conditions. For this study, what method is being used to control participant gender? |
|
Definition
|
|
Term
| What is the advantage of a laboratory study over a field study? |
|
Definition
| increased internal validity |
|
|
Term
| In a between-subjects design |
|
Definition
| each participant experiences only one level of the independent variable. |
|
|
Term
| An advantage of matching a variable across treatments rather than using random assignment to form your groups is that |
|
Definition
| matching guarantees that there is no systematic relationship between participant characteristics and the treatment conditions. |
|
|
Term
| What is the most common statistical analysis for a single-factor two-group design? |
|
Definition
| independent measure t test |
|
|
Term
| Compared to a between-subjects design, the variance in a within-subjects design is greatly reduced because the _______ are measured and statistically removed. |
|
Definition
|
|
Term
| A Latin square is used with |
|
Definition
|
|
Term
| Differential history effects are a threat to internal validity for which design? |
|
Definition
| pretest-posttest nonequivalent control group design |
|
|
Term
| The pre-post designs are similar to within-subjects designs, however in a pre-post design it is impossible to |
|
Definition
| counterbalance order of treatments. |
|
|
Term
| A researcher is examining motor skill development by observing children at 18 months old, 24 months old, and 30 months old. If the researcher uses a longitudinal design and obtains 20 scores for each age, how many children participated in the entire study? |
|
Definition
|
|
Term
| What is the correct description for a research study comparing problem solving ability for girls versus boys under three different levels of temperature? |
|
Definition
|
|
Term
| In a 2 x 3 between-subjects factorial experiment, there are _____ groups of participants |
|
Definition
|
|
Term
| What three outcomes are possible from a 2´2 factorial design? |
|
Definition
one main effect
no main effects
one interaction |
|
|
Term
| A Pearson correlation of r = +0.25 indicates that a graph of the data would show |
|
Definition
| points widely scattered around a line that slopes up to the right |
|
|
Term
| A researcher reports a positive relationship between sugar consumption and activity level for a group of 8-year-old children. However, the researcher cannot be sure whether the extra sugar is causing the children to be more active or whether the extra activity is causing the children to eat more sugar. This is an example of |
|
Definition
| the directionality problem |
|
|
Term
| The _______ method for quantifying observations involves counting the number of times a behavior occurs. |
|
Definition
|
|
Term
| Rosenham’s research investigating the experiences of mental patients and patient-staff interactions in psychiatric hospitals is an example of |
|
Definition
|
|
Term
| identify a potential problem with Internet surveys? |
|
Definition
| It can be difficult to control or even know the composition of the sample. |
|
|
Term
| A stable trend within a phase is defined as |
|
Definition
| a set of observations that cluster around a line sloping up to the right or a line sloping down to the right when graphed. |
|
|
Term
| Averaging over two or three consecutive observations is one method for dealing with |
|
Definition
|
|
Term
| The advantage of an ABAB design versus a simple AB phase-change design |
|
Definition
| the ABAB design reduces the probability that the result is just chance or coincidence |
|
|
Term
| What outcome is likely to occur for a hypothesis test evaluating a treatment that has a very large and robust effect? |
|
Definition
| correctly rejecting the null hypothesis |
|
|
Term
| Without some correction, measures of inter-rater reliability |
|
Definition
| tend to overestimate the true level of agreement between the raters. |
|
|
Term
|
Definition
| the average squared distance from the mean |
|
|
