final exam research methods
In a normal distribution, what percentage of scores falls between -1 and +1 standard deviations from the mean? Between the mean and -2 standard deviations? Between the mean and +3 standard deviations? Within the range of ±3 standard deviations?
+/- 1 → 68.26% +/- 2 → 95% +/- 3 → over 99%
Which is larger—a correlation of +.45 or a correlation of -.60? Explain.
-.60 because it is further from 0.
How many main effects can be tested in a 2 ´ 2 design? In a 3 ´ 3 design? In a 2 ´ 2 ´ 3 design?
-2 main effects -3 main effects
If you want to have 20 participants in each experimental condition, how many participants will you need for a 2 ´ 3 ´ 3 completely randomized factorial design? How many participants will you need for a 2 ´ 3 ´ 3 repeated measures factorial design?
-360 -20
What three criteria characterize good descriptions of data?
-Accuracy -Conciseness -Understandability
Discuss the relative advantages and disadvantages between within-subjects designs and between-subjects designs.
-Advantages of w/i subjects More powerful than a b/w subjects Can detect effects of the IV more easily Require fewer participants -Disadvantages of w/i subjects Order effects: participants behavior is affected by the order in which they participate in the various conditions of the experiment Practice effects: performance improves merely because they've been doing it Fatigue effects: participants get tired, bored, or less motivated Sensitization: participants begin to realize what the hypothesis is, as a result participants may respond differently
What is the most common type of survey research design?
-Cross-sectional survey design: single group of respondents is surveyed
Should demand characteristics be eliminated or strengthened in an experiment? Explain.
-Demand characteristics are aspects of a study that indicate to participants how they should behave -They should be eliminated, we do not want participants to react the way we want them to, we want them to act naturally and how they would normally act because that is true data
Why are psychologists sometimes interested in (a) demographic research?
-Describing and understanding patterns of basic life events and experiences such as birth, marriage, divorce, employment, migration, and death. -They are interested in the psychological processes that underlie major life events.
How do researchers detect and eliminate placebo effects?
-Detected by using a placebo control group and a true control group -Placebo control group receive inactive substance and true control group will receive no pill and no medicine
What are experimenter expectancy effects, and how do researchers minimize them?
-Experimenter expectancy effects when the experimenters expectations can distort the results of an experiment by affecting how they interpret the participant's data -By using a double blind procedure, both don't know what experimental condition the participant is in at the time
Under what conditions is a grouped frequency distribution more useful as a means of describing a set of scores than a simple frequency distribution? Why do researchers often add relative frequencies to their tables?
-Grouped frequency is more useful when there are many possible scores -Relative frequency of each class is the proportion of percentage of the total number of scores that falls in each class interval
Assuming that all confounds were eliminated, the means of the conditions in an experiment may differ from one another for two reasons. What are they?
-If the independent variable did not have an effect -Error variance
How does the longitudinal survey design help researchers to draw clearer conclusions about changes over time? What problem arises when respondents drop out of a longitudinal study?
-In a longitudinal survey design helps researchers look at changes in behavior over time because the respondents are surveyed more than once over time. -When respondents drop out, the sample is changed so we don't know if changes are because of the data collected over time or because changes in the sample
A successive independent samples survey design is used to examine changes in attitudes or behaviors over time, but results from such designs are often difficult to interpret. Describe the successive independent samples survey design and discuss why it is sometimes difficult to draw clear conclusions about the changes in attitudes or behavior that are observed.
-In successive independent samples survey design, two or more samples of respondents answer the same questions at different points in time. -Difficult because people are not the same. Scores reflect different samples of students
What is a main effect?
-Main effect is the effect of a single IV in a factorial design -Reflects the effect of a particular IV while ignoring the effects of the other IV's
What is a multiple baseline design, and when are such designs typically used?
-Multiple baseline design -Two or more behaviors are studied simultaneously -Used we need to know if the obtained effects are due to the IV as opposed to extraneous factors -Measuring several variables
Distinguish between the null hypothesis and the experimental hypothesis.
-Null hypothesis states that the independent variable did not have an effect on the dependent variable -Experimental hypothesis states that the independent variable did have an effect
Distinguish between an independent variable and a participant (or subject) variable.
-Participant variable reflect existing characteristics of the participants, non-manipulated variables -IV is manipulated variable
Tell whether each of the following relationships reflects a positive or a negative correlation: 1. The amount of stress in people's lives and the number of colds they get in the winter 2. The amount of time that people spend sun tanning and a dermatological index of skin damage due to ultraviolet rays 3. Happiness and suicidal thoughts 4. Blood pressure and a person's general level of hostility 5. The number of times that a rat has run a maze and the time it takes to run it again.
-Positive -Positive -Negative -Positive -Negative
How does descriptive research differ from other kinds of research strategies, such as correlational, experimental, and quasi-experimental research?
-Purpose is to describe the characteristics or behaviors of a given population in a systematic and accurate fashion -Examples: Survey, demographic, epidemiological research are examples -Is not typically designed to test a hypothesis, but rather to provide information about the physical, social, behavioral, economic, or psychological characteristics of some group of people
What is a restricted range, and what effect does it have on correlation coefficients? How would you detect and correct a restricted range?
-Restricted range is when you make the sample restricted to only one subset -It can cause the researcher to be misled into concluding that the two variables are only weakly correlated, if at all -You need to include a variety of subsets and not just one
Imagine that you were conducting an experiment to examine the effect of generous role models on children's willingness to share toys with another child. Explain how you would use (a) simple random assignment and (b) matched random assignment to equalize your groups at the start of this study.
-Simple random: get a list of all of the people and then use a computer generator and randomly assign groups -Matched random: give them a pretest on prosocial behaviors, then use that pretest to randomly assign them into groups
Can single-case experimental designs be used to test the effects of various levels of an independent variable (as in a one-way group design)? Explain.
-Single case experimental designs that present varying nonzero levels of the IV -Researchers continue to manipulate the IV by introducing new levels of it, returning to the baseline each time -A-B-BC-A-BC design A-baseline B-graduate exposure C-reinforcement =So yes
Why are psychologists sometimes interested in (b) epidemiological research?
-Study the occurrence of disease and death in different groups of people -Psychologists are interested for two reasons - Epidemiological data can provide info regarding groups that are at risk of illness or injury, thereby helping health psychologists target certain groups for interventions that might reduce their risk - Behavioral researchers are interested in documenting the occurrence of psychological problems and they conduct epidemiological studies to do so.
What is the rationale behind the ABA design?
-The most common single participant designs involve variations of ABA design -Demonstrate that an IV affects behavior by first showing that the variable causes a target behavior to occur, then showing that removal of the variable causes the behavior to cease (REVERSAL DESIGNS) -Baseline/control condition: participant is first observed in the absence of the IV -The IV is introduced and the behavior is observed again, we should see a change in behavior from the baseline to the treatment period
More specifically, what does a 95% confidence interval tells us?
-The probability that the population value of the mean falls within the confidence interval is .95 -95% of our confidence intervals would contain the mean of the population
Why do researchers use expericorr designs?
-They are used in an attempt to understand how certain personal characteristics relate to behavior under varying conditions -To look for differences in how male and female participants respond to an IV
How do mixed/expericorr designs differ from other experimental designs?
-They include both independent variables and participant variables and basically theyre designed to test if a particular IV had an effect on a particular groups characteristics
Why can't we infer causality from correlation?
-Variables can be strongly related, but that does not mean it causes it. To conclude that a variable causes another the three criteria that need to be met are - covariation -directionality -elimination of extraneous variables -Correlational research only satisfies the first one
How do researchers analyze the data from single-case experiments?
-Visual inspection/Graphic analysis -With graphs that show results individually for each participant -Resist analyzing their results in the forms of means, standard deviations, C.I. and other descriptive statistics based on group data
What does it indicate if a participant has a z-score of 2.5? -.80? .00?
-Z score describes a particular participant's score relative to the rest of the data -How far from the mean the participant's score falls in terms of standard deviations
The correlation between self-esteem and shyness is -.50, and the correlation between self-consciousness and shyness is +.25. How much stronger is the first relationship than the second? (Be careful on this one.)
.25
The correlation between self-esteem and shyness is -.50. Interpret this correlation.
.50% of shyness is attributed to self-esteem
What would it mean if eta2 or omega2 in an experiment was .25? .40? .00? What would it mean if d was .25? .40? .00? What would it mean if the odds ratio was .25? .40? .00?
1. .25 is a very weak relationship 2. .40 is a medium relationship 3. .00 is no relationship 4. if d was .25 is 25% of the SD 5. .40 is 40% of SD 6. and .00 is no % of the SD 7. Odd ratio is .25 is odds of on group are ¼ the odds in the other group 8. Odd ratio is .40 means the odds of one group is .4 the odds in the other group 9. Odd ratio is .00 is the odds of one group is .00 the odds in the other group
Why does quasi-experimentation sometimes require the use of "patched-up" designs?
1. Because it provides the most meaningful and convincing data possible 2. It helps because they do not have enough control over the environment to structure the research setting as precisely as they would like
What criticisms do proponents of single-case experimental designs level against group designs?
1. Error variance -Influence of unidentified factors that affect participants responses in an unsystematic fashion 2. Generalizability -Not generalizable 3. Reliability -Do not show if they are reliable or not
Discuss the trade-off between internal and external validity. Which is more important? Explain.
1. External: generalizability of research results 2. Higher internal validity leads to lower external validity, vice versa 3. Internal validity is more important
What are generational (or cohort) effects, and why do they sometimes create a problem in cross-sectional designs? How does the cross-sequential cohort design help to solve this problem?
1. Generational effects, people of different ages differ in the conditions under which their generation grew up 2. In cross sequential, two or more cohorts are measured at two or more times which helps separate age related effects from cohort and history effects
What is the difference between a histogram and a bar graph?
1. Histogram -Used when the variable on the x axis is on an interval or ratio scale of measurement. Bars on the graph touch 2. Bar graph -Bars do not touch
Why is effect size important? What are the three basic types of effect size indicators?
1. How strong of an effect the independent variable had on the dependent variable 2. Pearson correlation (r^2) Eta squared, omega squared 3. Cohen's d 4. The Odds Ratio
How do quasi-experimental designs differ from true experiments?
1. If the researcher lacks control over the assignment of participants to conditions and/or does not manipulate the causal variable of interest it is a quasi-experimental design 2. Unable or unwilling to manipulate the IV
What is the difference between interparticipant and intraparticipant variance? Which of these types of variance is more closely related to error variance in group experimental designs? Which type is of primary interest to researchers who conduct single-case experiments?
1. Interparticipant -Studying the effects on the independent variable on more than one participant 2. Intraparticipant -Replicated the effects of the independent variable with a single participant 3. Interparticipant is most closely related to error variance in group designs 4. Intraparticipant is of primary interest to researchers who conduct single case experiments
The text discusses several problems that are associated with null hypothesis significance testing. What are they, and why is each a problem?
1. It is based on an artificial dichotomy between "rejecting" or "failing to reject" the null hypothesis and the resulting decision to declare a particular finding "significant" or "non-significant" -The true likelihood of making a type I error ranges from .00 to 1.00, but we have created an arbitrary cutoff (usually at .05) at which we decide that a particular effect is statistically significant 2. Because journals generally publish only studies that find statistically significant effects, the research literature is based on a somewhat biased collection of results—those that met the magic .05 criterion -Assuming that the studies were competently conducted, those other unpublished, "nonsignificant" findings are relevant to a full understanding of the relationships between various independent and dependent variables. Failing to publish null results interferes with our ability to assess whether findings replicate and undermines the cumulative nature of science in which confirmations and disconfirmations should be considered together. 3. Because statistical significance is needed for publication, researchers have sometimes overanalyzed their data in search of significance (p-hacking) 4. The information that NHST provides is not as precise and informative as other approaches. -Only provides the means of the condition and the decision of whether or not to reject the null hypothesis. Means are just estimates of the entire population and the test only provides a dichotomous decision
Distinguish between a longitudinal design and a cross-sectional design.
1. Longitudinal -The IV is time -Same people, data is taken during multiple instances across time 2. Cross-sectional -Compare groups of different ages at a single point in time -Cannot distinguish b/w generational effects
Distinguish between the median, mode, and mean.
1. Mean Average, central tendency 2. Median Middle score of distribution 3. Mode Most frequent score
What is a dichotomous variable? What correlations are used for dichotomous variables?
1. Measured on a nominal scale but only has two levels (handedness - left or right) 2. Phi coefficient 3. A point biserial correlation (one is dichotomous other is interval)
Is the use of single-case studies a new approach to research in psychology? Explain.
1. No. They exist, they were used and preferred in the early days of psychology - many founders of behavioral science relied heavily on it 2. Today we use it for the study of operant conditioning
Distinguish between the nomothetic and idiographic approaches to behavioral science.
1. Nomothetic approach -Seeking to establish general principles and broad generalizations that apply across most individuals 2. Idiographic approach -Seeks to describe, analyze, and compare the behavior of individual participant
What are order effects, and how does counterbalancing help us deal with them?
1. Participants behavior is affected by the order in which they participate in the various conditions of the experiment -Practice -Fatigue -Sensitization 2. Counterbalancing is presenting the levels of the IV in different orders to different participants
What were some of Donald Campbell's contributions to behavioral research?
1. Popularized use of quasi-experimental designs 2. First to make distinction b/w external and internal validity 3. Unobtrusive measures 4. Interviewing techniques 5. Philosophy of science
What are the relative advantages and disadvantages of posttest-only versus pretest-posttest experimental designs?
1. Posttest only (DV is measured only after experimental manipulation) -Advantages -Capable of identifying effects of the IV and most experiments use them 2. Pretest-posttest -Advantages -Researcher can verify participants in the various experimental conditions did not differ w/ respect to the DV at the beginning of the experiment -By comparing pretest and posttest scores on the DV, reserachers can see exactly how much the IV changes participants responses -Pretest-posttest designs are more likely than posttest only designs to detect the effects of the IV on the DV
What three criteria must be met to establish that one variable causes changes in behavior? Which of these criteria are met by quasi-experimental designs? Which of these criteria are not met, and why?
1. Presumed casual variable preceded the effect in time 2. The cause and the effect covary 3. All other alternative explanations of the results are eliminated through randomization or experimental control 1. Quasi-experimental designs meet the first two criteria, but not the third because we do not have control over extraneous variables
Describe how participants are assigned to conditions in randomized groups, matched-subjects, and repeated measures experimental designs.
1. Randomized groups -Participants are randomly assigned to one of two or more conditions 2. Matched-subjects -Participants are matched into blocks on the basis of a variable the researcher believes relevant to the experiment. Then participants in each matched block are randomly assigned to one of the experimental or control conditions 3. Repeated measures -Each participant serves in all experimental conditions
Explain the difference between rejecting and failing to reject the null hypothesis. In which case does a researcher conclude that the independent variable has an effect on the dependent variable?
1. Rejecting the null hypothesis -Means that the researcher concludes that the independent variable did have an effect 2. Failing to reject the null hypothesis -Means that the researcher concludes that the independent variable had no effect
Discuss how the interrupted time series design with a reversal and the control group interrupted time series design improve on the simple interrupted time series design.
1. Reversal -Observing what happens to participants behavior when the quasi-IV is first introduced and then removed -Would help us to see if the IV actually had an effect or if it was due to history effects 2. Control group -Adds comparison groups. By measuring more than one group on several occasions, only one of which receives the quasi-IV, we can examine the plausibility of certain alternative interpretations -Helps us to rule out certain history effects
In what areas have single-case experiments been primarily used?
1. Study of basic learning processes 2. Operant conditioning 3. Animal researchers 4. Demonstrational purposes to show that a particular behavioral effect can be obtained 5. Study the effects of behavior modification
What three rules govern the construction of a grouped frequency distribution?
1. The class intervals are mutually exclusive (people can't fall into more than one group) 2. The class intervals capture all possible responses 3. All class intervals are the same size
A well-designed experiment possesses what three characteristics?
1. The researcher must vary at least one IV to assess its effects on participants' responses 2. The researcher must have the power to assign participants to the various experimental conditions in a way that ensures their initial equivalence 3. The researcher must control all extraneous variables that may influence participants' responses
What are four primary reasons that behavioral researchers use case studies?
1. To provide a source of insights and ideas 2. To describe rare phenomena 3. To conduct psychobiographical research 4. To supplement empirical data with illustrative anecdotes
Distinguish among treatment, confound, and error variance.
1. Treatment -Portion of the variance in participants' scores on the dependent variable that is due to the independent variable 2. Confound -When a variable other than the IV differs between the groups, confound variance is produced -Secondary variance, the portion of the variance in participant's scores that is due to extraneous variables that differ systematically b/w experimental groups 3. Error variance -Unsystematic differences among participants -Participants differ at the time they enter the experiment
What can researchers do to minimize error variance?
1. Use homogenous sample 2. Aside from diff in the IV, treat all participants the same at all times 3. Hold lab conditions constant 4. Standardize all research procedures 5. Automate experiment as much as possible 6. Use only reliable measurement procedures
What is program evaluation? Why do program evaluators rely heavily on quasi-experimental designs in their work?
1. Uses behavioral research methods to assess the effects of interventions or programs designed to influence behavior 2. They use them because the programs are not under the researcher's control, so they use quasi-experimental to evaluate their effectiveness
What threats to internal validity are present when the nonequivalent control group posttest-only design is used? Which of these threats are eliminated by the pretest-posttest version of this design?
1. We have no way of knowing whether the two groups were actually similar before the quasi experimental group received the treatment 2. If the 2 groups differ when they are measured at time O we do not know whether the difference was caused by variable X or if they differed before the quasi-experimental group received X 3. We have no way of being sure that the groups were equivalent before the participants received the quasi-IV, the non equivalent control group posttest-only design is weak in terms of internal validity 4. The pretest-posttest lets us see whether the 2 groups scored similarly on the DV before the intro of the treatment at point X.
What is experimental contamination? Why does a concern with contamination sometimes lead researchers to use quasi-experimental designs?
1. When participants in different conditions of a study interact with one another, possibility exists that they may talk about the study among themselves and that one experimental condition becomes contaminated by the other 2. Because it minimizes contamination b/c it lessens the likelihood the people in the conditions will interact
Define the confounds in the following list and explain why each confound undermines the internal validity of an experiment: biased assignment of participants to condition differential attrition pretest sensitization history miscellaneous
1. biased assignment of participants to condition -results can be significant or insignificant due to biased assignment, not because there is actual effects 2. differential attrition -weakens internal validity, when the rate of attrition differs across the experimental conditions 3. pretest sensitization -participants react differently sometimes when they are pretested, the researcher may conclude that the IV has an effect when the effect is influenced by the pretest 4. history -history effects occur when an event unrelated to the study changes the results 5. miscellaneous design confounds -Design issues like not treating the participants the same can have an effect
How many conditions are there in the simplest possible experiment?
2
How many interactions can be tested in a 2 ´ 2 design? In a 3 ´ 3 design? In a 2 ´ 2 ´ 3 design?
2 ' 2 has one interaction 3 ' 3 has one interaction 2 ' 2 ' 3 has 3 interactions
How many independent variables are involved in a 3 ´ 3 factorial design? How many levels are there of each variable? How many experimental conditions are there? Draw the design.
2 IV's, each with 3 levels 9 conditions
How many levels of the independent variable are there in an ABACADA design?
3
Describe a 2 ´ 2 ´ 3 factorial design. How many independent variables are involved, and how many levels are there of each variable? How many experimental conditions are in a 2 ´ 2 ´ 3 factorial design?
3 IV's, 2 have 2 levels, 1 has 3 levels 12 conditions
What is a case study?
A detailed study of a single individual, group, or event
What does it mean if the difference between two means is statistically significant?
A statistically significant finding is one that has a low probability (usually <.05) of occurring as a result of error variance alone
What is raw data?
All participants' scores on all measures
Why can it be argued that the formula for calculating r should be named the Edgeworth, rather than the Pearson, correlation coefficient?
Because Francis Edgeworth technically discovered a very similar formula a few years earlier but didn't recognize the importance of his finding. When Karl Pearson derived the formula several years later, he did not know about Edgeworth's formula but since he was the one to popularize it and apply it to problems in biology and psychology, it became known as the Pearson correlation coefficient.
Why do researchers often examine scatter plots of their data when doing correlational research?
Because it allows them to see their entire sample plotted and the scatter plot allows them to see the relationship between their variables
Why may we not interpret or discuss a correlation coefficient that is not statistically significant?
Because it is of no value to us
Why do researchers calculate the coefficient of determination?
Because it makes it easier to interpret r (correlation coefficient)
Why do researchers prefer the standard deviation as a measure of variability over the range and the variance?
Because standard deviation takes into account all of the scores when calculating variability and it expresses behavioral variability in the original units of their data.
When a longitudinal design reveals a change in behavior over time, why can we not conclude that the change is due to development?
Because there are other factors that can induce the change
In what way do researchers take a risk if they do not pilot test the independent variable they plan to use in an experiment?
Because they do not ensure that the levels of the IV are different enough to be detected by the participant
When analyzing data, why do researchers test the null hypothesis rather than the experimental hypothesis?
Because we can only provide sufficient evidence for the null hypothesis, we can't prove anything else to be true or false
Why is contemporary history a threat to internal validity in a simple interrupted time series design?
Because we cannot rule out the possibility that the observed effects were due to another event that occurred at the same time as the quasi-IV
Why must researchers ensure that their experimental groups are roughly equivalent before manipulating the independent variable?
Because we want to ensure that the differences we see in the groups after manipulation of the IV was due to the IV, not because of difference in the groups
Which is worse—confound variance or error variance? Why?
Confound because we can't get rid of it and it invalidates an experiment whereas error variance does not
Do outliers increase or decrease the magnitude of correlation coefficients?
Decrease
What important information do researchers obtain from conducting a power analysis?
Determines the number of participants needed in order to detect the effect of a particular independent variable.
Explain how you would conduct the study in Question 11 as a within-subjects design.
Each participant must serve in all conditions
Give your own example of an environmental, instructional, and invasive experimental manipulation.
Environmental -Experimental modifications of aspects of the research setting -Researcher studying learning will vary the amount of reinforcement provided Instructional -Vary the IV through instructions or information that participants receive -In a study about test taking participants will be given different instructions - focus on amount correct or focus on amount answered Invasive -Involve creating physical changes in the participant's body through physical stimulation, surgery, or drugs -The effects of a medicine
How are error bars interpreted?
Error bars provide info about the researchers confidence in the value of each mean. They show where the C.I. lies (top to bottom) of the data
What advantage do experiments have over descriptive and correlational studies?
Experimental designs allow researchers to draw conclusions about cause and effect relationships and test the hypothesis
What is a factorial design? Why are factorial designs used more frequently than one-way designs?
Factorial design is when two or more IV's are manipulated -They provide info about the effects of each IV and the IV's when they're combined. -Allows us to identify individual effects, combined effects, and error variance
In general, what does a confidence interval tell us?
Gives us a point estimate of the population value
What does the confidence interval tell us?
How accurately the mean we calculate in our study estimate the population mean
How do we know whether a particular score is an outlier?
If it is further than 3 SD's away from the mean of the data
Under what circumstances would a researcher use a quasi-experimental rather than an experimental design?
If the researcher lacks control over the assignment of participants to conditions and/or does not manipulate the causal variable of interest it is a quasi-experimental design
What is a psychobiography?
Involves applying concepts and theories from psychology in an effort to understand the lives of famous people
How can partial correlation help researchers explore possible causal relationships among correlated variables?
It allows researchers to examine a third variable's possible influence on the correlation b/w two other variables → tests the correlation of x & y while removing the influence of variable z
Why should researchers never use the one-group pretest-posttest design?
It fails to eliminate most threats to internal validity
What is an interaction?
It is present when the effect of one IV differs across the levels of other IV's
What is the relationship between confounding and internal validity?
It screws with the validity of the study, higher confounding variance, lower validity
Explain the rationale behind time series designs.
Measure the DV on several occasions before and on several occasions after the quasi-V occurs. By measuring the target behavior on several occasions, researchers can see whether changes in the DV coincide precisely with the intro of the quasi-IV
Describe a mixed, or split-plot, factorial design. This design is a hybrid of what two other designs?
Mixed -Design that combines one or more between subjects variables with one or more within subjects variables. -Randomized and repeated measures design
Must all experiments include a control group? Explain.
No, sometimes there is no need or use for them
Does a nonequivalent groups pretest-posttest design eliminate local history as a potential explanation of the results? Explain.
No, we don't know the histories of the people in the group, so we cannot control for that
What are some pros and cons of conducting descriptive research using the Internet?
Pros -Inexpensive -Cost and time of data entry is lowered -Reduces mistakes -Allows researchers to contact respondents who would otherwise be difficult to reach (allows wide array of people to be contacted, no geographical barriers) -Respondents can reply at their convenience Cons -Researcher has little control over the selection of sample -People without internet cannot participate (lower incomes, less educated, rural areas, over age of 65 are underrepresented) -Probability samples can't be used -Difficult to verify that respondents are who they say they are (underage) -People can respond to survey more than once (especially if they are being paid)
Distinguish between qualitative and quantitative levels of an independent variable.
Qualitative IV -Participants are treated in different ways Quantitative IV -Four levels of the IV, 0 100 300 or 600 mg of caffeine
Distinguish among randomized groups, matched-subjects, and repeated measures factorial designs.
Randomized groups -Participants are assigned randomly to one of the possible combinations of the IV Matched subjects -First matching participants into blocks on the basis of some variable that correlates with the DV, then they are randomly assigned in each block Repeated measures -Participants participate in every experimental condition
Explain what it means if a researcher sets the alpha-level for a statistical test at .05.
Researchers only reject the null hypothesis if there is less than a .05 chance that the difference they obtain between the means of the experimental groups is due to error variance rather than to the independent variable
Researchers who use group designs replicate their findings by repeating the same (or a similar) experiment on other samples of participants. How do single-case researchers replicate their findings?
Single case designs allow the generality of one's hypothesis to be assessed through replication on a case by case basis
Draw a negatively skewed distribution.
Tail goes to the left
Describe the simple interrupted time series design.
Taking several pretest measures before introducing the IV and then taking several posttest measures afterwards
Explain how you would use a manipulation check to determine whether you successfully manipulated room temperature in a study of temperature and aggression.
Test different room temperatures on a small group of participants
What does a coefficient of determination of .40 indicate?
That the variable is only .40 correlated, it is a weak correlation
What impact does reliability have on correlation?
The less reliable our measures, the lower the correlation coefficients we will obtain
What does a p-value tell us?
The probability that the findings of the Ho can be rejected or fail to be rejected
Why do researchers use null hypothesis significance testing?
Used to determine whether differences b/w the means of the experimental conditions are greater than expected on the basis of error variance alone. Because error variance can cause the means of the conditions to differ even when the IV had no effect, we need a way to estimate the probability that the effect we obtained is duet o error variance versus the IV and significance testing provides this estimate -Researchers use of significance testing to determine the probability that an obtained effect is a real effect of the IV or simply the result of random error variance
When is the Spearman rank-order correlation used?
When one or both variables are measured on an ordinal scale
Under what circumstances is an ABA design relatively useless as a way of testing the effects of an independent variable?
When the IV produces permanent changes in a participants' behavior, changes that do not reverse when the IV is removed
Under what conditions is the median a more meaningful measure of central tendency than the mean?
When there are outliers