[M] Experimental Methods in Psychology
[M] Experimental Methods in Psychology Psych 312
Popular in Course
Popular in Psychlogy
This 21 page Study Guide was uploaded by Lazaro Feest on Thursday September 17, 2015. The Study Guide belongs to Psych 312 at Washington State University taught by Lisa Fournier in Fall. Since its upload, it has received 71 views. For similar materials see /class/206007/psych-312-washington-state-university in Psychlogy at Washington State University.
Reviews for [M] Experimental Methods in Psychology
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/17/15
Exam 2 Review Psychology 312 Dr Fournier You will be responsible for all information presented in lecture lab and in the textbook Ray Ch 8 9 10 11 and 14 Exception lab research paper background readings See Dr Fournier or you re your lab instructor if questions 0 Sampling different types be able to identify when might use a certain type discussed in Ch 13 0 Probability sampling equal chance that any 1 individual is selected from the population you wish to generalize to makes it random I Simple Random Sampling each member has an equal chance of being selected draw names from hat Systematic Sampling list of total population and choose every 11th person listed list order should be unbiased alphabetical list Strati ed Random Sampling When you wantneed certain types of people to be systematically represented in the sample divide population according to the category eg minority and take a separate random sample from each of the resulting divisions 50 whites 50 blacks 50 hispanics Cluster Sampling randomly select a certain number of population units like city blocks and then enlist people within those population units Multistage Sampling randomly sample at different stages in the process randomly sample population units THEN randomly sample within these units is used when the population is larger randomly choose colleges then randomly sample students 0 Non probability sampling not random I Convenience Sampling use individuals who are available to us ndings may not be generalized psych students Quota Sampling set up a quota or number of speci c types of individuals 3 athletes 3 frat boys 3 honor students Snowball Sampling use when there is no list of the population locate 1 person and get more of sample from contact with this person 0 Survey questionnaire research 0 Format types of questions pros and cons of each general considerations I Open ended no xed answer freedom to respond in any manner 0 Advantages 0 Doesn t impose researcher s point of view 0 P can provide info researcher had not considered 0 Disadvantages o Dif th to analyze and establish meaningful patterns 0 Dif th to translate answers into categories that t the hypothesis I Fixed Alternative P picks a response listed 0 Advantages 0 Easy to translate into categories 0 Easy to analyze o Disadvantages 0 Limits number of responses 0 Can t give reason for response I Combined Open and Closed Ended can pick a xed response or another that is not listed 0 Advantages 0 Same as those of open and xed alternative 0 Disadvantages o Open ended may be dif th to classify 0 Use xed answers as a guide I Funneling being with open ended question and follow it with more speci c response alternatives 0 Advantages 0 Get them to think about answer and how it ts into speci c alternative before they select the alternative 0 Disadvantages 0 Don t know why they chose a particular response I Creating Response Alternatives 0 If asking about frequency of a behavior 0 Never rarely sometimes often very often 0 If asking whether something is true 0 Very true somewhat true not very true not at all true 0 Make sure categories do no overlap 0 Methods of administering different types pros and cons of each I Liker used to measure attitude 0 Assign response to a number I Semantic differential uses bipolar adjectives good and bad 0 Adjectives placed at ends of 7 point scale 0 What factors in uence your sample size I How many people are available for your research and how homogeneousheterogeneous is this group 0 If homo on characteristics we are studying we can use fewer o If hetero on characteristic we need more subjects I How much difference can the researcher tolerate between sample and overall population studying 0 Larger population bigger sample 0 Greater variability in population bigger sample I Number in sample can be based on previous similar research 0 Better to err on larger vs smaller sample size 0 What types of conclusions can you draw based on this type of research I Correlational 0 Know how to plot data if given data know how to appropriately plot it know how to label axis related to IV amp DV I IV on x axis 0 One factor along x axis other factor signi ed by differently shaded bars in a bar graph I DV on y axis Know how we can increase the power of our experiment 0 Larger sample size more power Know steps of scienti c research in general and why this is a circular process Controliwhy is control important Be able to identify problems related failure in implementing control 0 Control related to participant assignment amp selection who how many how select how we will assign to groups Random assignment to groups random selection how affects within group and between group variance I Control ensures experimental results are due to the IV and not any other uncontrolled factors internal validity I Random Assignment to conditions 0 Controls for knownunknown confounding variables 0 Increases internal validity 0 Any suspectedunsuspected selection biases will be nulli ed across conditions 0 DOES NOT REMOVE DIFFERENCES AMONG PARTICIPANTS differences are equated across conditions 0 Control achieved through experimental design pretestposttest Solomon 4 group etc I Design must perform 2 functions 0 Allow evaluation of the effects of the IV on the DV 0 Help rule out confounds I Pretest Posttest control group design Randomly assign Ps to groups Give pretest to each group before treatment Give posttest to each group after treatment Benefrts controls several factors GIVES US AN IDEA IF THE 2 GROUPS ARE EQUAL BEFORE THE TREATMENT However the pretest could in uence the outcome of the experiment if so use the Solomon Four Group Design 1 Pretest Treatment Posttest 4 Posttest If the pretest didn t in uence treatment group then group 1 and 3 would have the same posttest If the pretest didn t in uence the control group then group 2 and 4 would have the same posttest 0 Compare group 1 and 2 to nd the treatment effect Control related to the logic of experimentationi4 steps participant selection participant assignment design amp consistency of treatment interpretation of hypothesis see lecture notes I If control is achieved through participant assignment or design consistency of treatment you can assume that the results are due to the IV Between subject designs within subject designs 7 be able to identify de ne know differences between when might use 1 design over the other advantagesdisadvantages of each these types of designs 0 Between Subjects Design I Any given individual or group does not receive all levels of the IV 0 Within Subjects Design I All levels of the IV are given to an individual or group Mixed subject designs 0 At least 1 IV is run between subjects and at least 1 IV is run within subjects Types of between subject designs completely randomized design multileveled completely randomized design 0 Completely Randomized I 1 IV with 2 levels 0 Subjects are randomly assigned to either a control or treatment groups 0 Multi Level Completely Randomized I 1 IV with 3 or more levels 0 Random assignment of subjects into one of the groups one for each level of the IV Factorial designs within amp between subjects 7 de ne be able to identify IVs factors and levels of each IV be able to identify the type of factorial design e g 2 X 2 2 X 3 2 X 2 X 2 etc Be able to answer the following related to factorial designs 0 What do factors represent I Independent variables 0 How identify of IVs I 2x3x3ORAxBxC3factors o How identify levels of each factor I The numbers in the above example 0 How many conditions Multi 1 the number of levels 0 I One How many possible main effects might we have I Equal to the number of factors How many possible interactions might we have I How many combinations of factors we have I 2 x 3 1 I 2x3x34 Be able to plot data correctly and identify main effects and interactions 0 O O 0 Be able to de ne and determine whether a main effect exists I Make a table compare means 0 Be able to de ne and determine whether an interaction effect exists Make a table com are the differences I Interaction effects 0 What is the power or advantages of factorial designs over nonfactorial designs I Can examine more than one hypothesis or factor 0 Behavior is often in uenced by more than one factor I Can simultaneously evaluate each factor 0 More economical I Can examine interactions between factors Given a scenario be able to identify the design in terms of the factors Be able to identify hypothesis independent and dependent variables Be able to identify main effects and interactions Be able to interpret main effects and interactions will need to describe based on data from a scenario Know the 8 possible data outcomes of a 2 X 2 factorial experiment see chapter 8 and know how to interpret the data main effects interactions presented in each of the possible data outcomes Know advantagesdisadvantages of betweensubject and withinsubject designs 0 Between subject I Advantages 0 Each participants receives only one level of the IV 0 No fatigue or order effects 0 Can be more certain that the IV caused the difference I Disadvantages 0 Need more participants 0 Within subject I Advantages o Equate groups 0 Reduce the number of participants needed 0 Statistically more sensitive to changes in the treatment effect 0 Reduces error VAR o Increases sensitivity of study 0 F Ratio error VAR is statistically smaller 0 Smaller treatment differences allow us to reject the H0 0 I Disadvantages 0 Not appropriate when 0 Treatment has a lasting effect exposure to one IV might carry over into another conditions 0 Extremely sensitive to time related effects confounds experience in one condition can in uence performance in another Be able to identify main effects amp interactions and interpret these see lab amp class exercises Most likely you will be given an experimental scenario a table of data and or a graphical representation of main effects andor interactions and asked to interpret what they mean in terms of the independent amp dependent variables and in terms of the hypothesis There are many questions concerning interpretation of results main effectsinteractions Be able to identify know pros amp cons of withinsubject and betweensubject design when one of these designs may be more appropriate Withinsubject designs in general and how affects within group variance significance tests 0 Order effects and sequence effects I Experience in 1 condition can in uence performance in another 0 Issues related to counterbalancingifull counterbalancing partial counterbalancing Know how to figure out how many participants need to fully counterbalance or partially counterbalance conditions in study Know K Know how to due a Latin Square Counterbalancing Design I Complete Counterbalancing o All conditions in experiment appear in all possible orders 0 Every possible sequence appears at each presentation of treatment 0 Use K to determine all possible orders required to completely counter balance I Incomplete Counterbalancing o Counterbalance with selected orders 0 Each conditions in all possible orders BUT each condition does not follow each of the other conditions 0 Use Latin Square to determine the conditions I K 0 K Number of conditions 0 KK 1K 2K 3 o If there are 6 conditions I 66x5x4x3x2x1howmanyparticipants needed 720 I Latin Square 0 Equation 1 2 N 3 N 1 4 N 2 5 N 3 5 N 4 o N of conditions 0 If there are 9 conditions you need 9 columns 0 129384756 1 2 9 3 8 4 7 5 6 2 3 l 4 9 5 8 6 7 3 4 2 5 l 6 9 7 8 4 5 3 6 2 7 l 8 9 5 6 4 7 3 8 2 9 l 6 7 5 8 4 9 3 l 2 7 8 6 9 5 1 4 2 3 1 6 2 9 1 8 2 7 3 6 4 5 o Matchedsubject procedures 0 We use these when you cannot use a within subjects design due to carry over effects practicefatigue effects or to overcome any pitfalls with within subjects designs as a control procedure I Match pair of individuals on a variable that is correlated with the DV and run experiment wo analyzing the matched variable as an experimental procedure I Individuals are matched on a factor then run the experiment and analyze that factor Ecology and factors that affect internal amp external validity Be able to identify factors below and how to avoid such factors 0 experimenter factors I interrater reliability 0 degree of agreement among raters gives a score of how much consensus there is in the ratings helps to nd outliers and common scores and validates the experiment 0 Standardize instructions 0 Researcher blind to hypothesis and conditions 0 Participants should always be blind to the hypothesis 0 Double blind study I experimenter bias 0 Changes in the DV produced by the attitude or behavior of the researcher not related to the IV 0 Expectations about the outcome of the study in uences the results obtained 0 Be blind to the hypothesis and conditions 0 Use more than one rater 0 Look at inter rater reliability oo o 1 u U 4 O O 0 subject factors I Hawthome effect 0 Participants think they are special when they are in a study 0 They try harder than usual 0 Is a threat to ecological validity 0 Use double blind study I placebo factors 0 Suggestion of change can produce a change 0 Use double blind study I demand characteristics 0 Participant is more in uences by the experimental setting than by the IV 0 Use double blind study and have no leading suggestions I cultural and social bias 0 Social desirability bias want to give socially appropriate answers Response set response is unrelated to the question agree w everything or disagree w everything Emphasize anonymity Use questions that may less likely suggest moral judgment is involved 0 Use similar questions with different answers 0 doubleblind experiment I Participant and experimenter are unaware of the condition the participant is assigned to You will be given a few experimental scenarios and will be asked questions related to any of the information above You will need to know all of the material above at a conceptual level The test will be mainly multiple choice some short answers and essays Don t stress Just study Know information from text lectures and lab activities If questions come see the instructor or your lab TA Exam 1 Review 2 pages Psychology 312 Dr Fournier You will be responsible for all information presented in lecture lab and in the textbook Ray Ch 1 2 3 5 6 and 7 Come see me during office hours if questions Why is understanding research methods important 0 Methods will help you explore understand and interpret various phenomena that you are curious about 0 As a consumercitizen you will often need to evaluate and act upon research results 0 To find answers to such questions you must rely on data and interpretation of data provided by experts o Experts can be wrong 5 sources of knowledge and how they are related to scientific research 0 Tenacity Acceptance of a belief wo data to support it 0 Common Sense Accept ideas based on past experiences or perceptions 0 Authority Accept as true info because it is communicated by experts 0 Reason Accept ideas based on the assumptions that the rules of logic will permit only correct conclusions to follow from premises 0 Science Accept ideas based on objective observation and reason Assume that events have finite causes causeeffect relationships are stable over time and there are no nonnaturalsupernatural causes Why is experimental psychology considered to be an empirical science Conclusions are based on direct observations not speculation traditional beliefs common sense or reasoning alone Why is Experimental psychology considered to be theoretically driven like other sciences 0 We don t 39ust collect facts we exlain what is observed usin theories l r 0 How does science progress 0 There are often a couple oftheories to answer a particular question with science we test which theory is best then go off ofthat theory Why do vague theories hinder the progress of science o It is hard to test a vague theory therefore hard to prove it wrong or improve upon it Know basics of Experimental research Vs Correlational research naturalistic observation case studies survey research 0 Experimental direct manipulation of variables to allow examination of causeeffect relationships 0 Correlational ask whether there is an association between two variables but does not establish how these two variables influence each other Know how to set up a naturalistic observation when might use and characteristics of o Noninterference observe naturalreal patterns in the world qualitative information that only gives description 0 You watch without interacting with the subject and without the subject knowing that you are watching Know what it means when 2 variables are correlated or not correlated Be able to identify positive negative and no correlation Know the 3rd variable problem 0 Positive corr as one increases so does the other 0 Negative corr as one increases the other decreases o No corr they do not move together one increasing does not mean the other will increase or decrease Inference know what term means amp how applied lnferences are related to the hypothesis or some larger theory that we want to evaluate make inferences about the research participants behavior from what we observe Construct validity know what term means amp how applied s our measure accurately reflect your construct what we are trying to describe internal amp external validity know what term means amp how applied o Is the experiment internally valid Are other variables besides the N that could have lead to our findings Are there confounds Are there extraneous variables e experiment externally valid eneralized to the ulation l Operational definitions relate to hypothesis IV and DV be able to translate IV and DV into Be able to take a construct and operationally define 0 An objective definition of a construct stated in terms of how one will measure that construct we try to use more than one operational definition 0 Ex The construct is anxiety the operational definition could be physiological measures like heart rate a psychometric scale of anxiety or a self report lnductive vs deductive reasoning know each how different which most popular in science strengths and weaknesses of approaches when one may be used vs other lnductive when you go from data results from a study to a general theory premise Makes generalizations from a specific instance or several new facts to a general idea Weakness may overlook unobserved factors that may be responsible for the effects we observe Strength good starting point when little is known about the phenomenon we wish to study 0 Deductive when you begin with a general theory premise and lead to particular statements concerning data conclusions Scientists use this lfThen logic Weakness could use invalid reasoning Strength allows us to draw conclusions Deductive reasoning valid confirmation invalid confirmation valid disconfirmation and invalid disconfirmation be able to identify each make sure you do not use either type of invalid reasoning above when drawing conclusions from an example scenario Which type of reasoning listed above is the one science uses or should use 0 Valid Confirmation orrect conclusion If P then Q P occurs therefore we should see Q o Invalid Confirmation ncorrect conclusion If P then Q Q occurs therefore we should see P o This does not make sense 0 Valid Disconfirmation this is the most powerful reasoning in science To disconfirm present theory If P then Q Don t observe Q therefore should not observe P o Invalid Disconfirmation To incorrectly disconfirm a theory If P then Q Don t observe P therefore should not observe Q o This does not make sense hypothesis general and specific based on operational definitions be able to construct and identify 0 Come up with an ideaexpectation called the hypothesis evaluate the hypothesis through observation and research and draw conclusions about the hypothesis 0 What will you manipulate Measure Predict Evaluate IV and DV IV what the experimenter directly manipulates DV what the experiment measures compare levels of IV 0 Evaluate Treatment Effect The difference in magnitude ofthe DV between the control and treatment conditions 0 Make hypothesis concrete Use operational definitions Construct validity 0 Make predictions Indicate predicted outcome based on the IV and DV independent and dependent variable make sure can operationally define identify and create IV amp DV for different scenarios Alternative hypothesis vs Null hypothesis what each means Validity and reliability know what mean and be able to identify or come up with an example can have one without the other 0 Validity accurate must reflect the variable you measure 0 Reliability consistent must show an effect repeatedly Scaling all different types of scaling how to identify which type of scaling is appropriate what kinds of conclusions you can draw with different types of scaling Given a scenario should be able to answer these questions Identify incorrect conclusions based on misinterpretations of scaling O 0 Nominal nonordered categories differences between categories is qualitative used for measuring race etc kes no sense to add subtract multiply or divide o Ordinal ordered categories numbers reflect quantitative difference no absolute zero cannot say one value is twice that of another ex rank sports teams Differences between categories are not equal intervals 0 Interval ordered categories with equal intervals no absolute zero cannot say something is twice the value of another ex temp in F 0 Ratio ordered categories with equal intervals and an absolute zero Statistics mean median mode what mean how calculate how related to distribution of scores Mean median and mode the same normal distribution Mean greater than the median positively skewed Mode greater than the mean negatively skewed normal distributions bimodal skewed importance of variability know what means in terms of significant differences between sample means Helps determine ifthe means between the 2 conditions are truly different 0 Zscores when use amp what can tell us zscore for one particular measurement is a deviation score that is calculated by subtracting the mean ofthe distribution from that score and then dividing the difference by the standard deviation Relates all the scores to the mean can compare scores easily Correlation what conclusions can draw interpretation of positive and negative correlation Know from example scenario whether demonstrates positive negative or no correlation Types of plots when plotting data when 1 is more appropriate than another scatter diagram bar graph line graph Scatter Diagram Bar Graph Line Graph Know how to plot data if given data know how to appropriately plot it know how to label axis related to IV amp DV IV y axis 00000 O O O axrs What type of research design allows us to make causeeffect relationships Why Be able to explain 0 Experimental because we can control variables and test to see what variable causes a difference Differences between controlled experiment correlation amp naturalistic observation in terms of methods conclusions that can be drawn amp circumstances most likely use Be able to identify whether a measure is reliable andor valid know definition of both ofthese terms lnferential statistics know what allows you to do assumptions what is the purpose Know what we assume conceptually when we test for significant differences between groups using inferential statistics and know statistically what we assume when we do inferential statistics see lecture amp Ch 5 0 When we attempt to understand the properties of all possible samples called the population by looking at the particular sample with which we are workin Assume that the sample we are discussing is the result of random sampling or some systematic form for sampling 0 Using sample means to infer things about the population mean What is alpha level Beta level 1alpha 1B Why do we have an alpha level What is the p level Is there such thing as a p 0 level of significance Why or why not Do we want our p level to be high or low Why 0 Alpha Level the probability that you will commit a Type Error usually set at 05 meaning there is a 5 5 out of every 100 results are due to chance p 0 does not exist because there is always the possibility that your results are due to chance Want p to be low 0 Beta Level the probability that you will commit a Type II Error want power to be high and beta to be small In general how does probability relate to inferential statistics 0 Used to determine ifthe results are generalizeable to the population Null Hypothesis Ho alternative hypothesis H1 or HA operational definition treatment effect treatment and control group O Evaluation ofthe research hypothesis alternative hypothesis 3 questions need to ask see lecture notes amp Ch 6 Internal and external validity 0 Internal Validity the ability to make valid inferences concerning the relationship between a dependent and independent variable in an experimental situation 0 External Validity generalizability of an experimental outcome to other groups settings treatment variables and measurement variables Confounds extraneous variables know how to identify each know how each affects variance know how each effects the probability of rejecting the null hypothesis know all differences between confounds and extraneous variables 0 Confound systematically affects the results affects betweengroup variance and causes Type Error can lower between group variance increases chances of rejecting null o Extraneous variables affects within group variance causes Type II Error can higher rou variance decreases chances of re39ec 0 null Know how to identify and control for any extraneous variables Withingroup variance betweengroup variance know what each is identify under which conditions each type of variance may increase which type of variance do you want to be high Which low Know how confounds and extraneous variables affect each type of variance know what you can do to increase your chances of supporting your alternative hypothesis eg find significant differences between your groups and know how each type of variance affects the probability of rejecting the null hypothesis Know chance variation and systematic variation and how these types of variation map on to within group and between group variance When is systematic variation good When is it bad What is the F ratio based on What does an F 1 mean Do you want F to be high or low 0 Betweengroup variance divided by withingroup variance 0 Want it to be greater than one o F 1 means there is no effect Know what distributions look like for high withingroup variance low withingroup variance high betweengroup variance low betweengroup variance Be ready to compare distributions that may be high or low in terms of one type of variance or both Know type and type II error Know definitions of both which type of error is worse and why how to avoid these types of errors be able to identify if a type or type II error in a scenario know each type of error in terms of true state of affairs Know when we can feel confident that our results are due to our lV Know what a correct rejection and correct acceptance ofthe null hypothesis is Do we ever seekto support the null hypothesis Why or why not Know alpha level in relation to null hypothesis Know how we can increase the power of our experiment Know steps of scientific research in general and why this is a circular process 0 1 Participant Selection 0 2 Participant Assignment 0 3 Design and Consistency of Treatment orm experiment 0 4 Interpretation of Hypothesis Null hypothesis Confound hypothesis Research hypothesis 0 You might find support for a hypothesis then go back and refine it and start over You will be given a few experimental scenarios and will be asked questions related to any of the information above You will need to know all of the material above at a conceptual level The test will be mainly multiple choice and an essay question Don t stress Just study Know information from text lectures and lab activities If you have questions make an appointment with the instructor or your lab TA or discuss with TA during lab Exam 3 Review Psych 312 Dr Fournier Questions will be based on lectures class activities lab activities and readings from the textbook Questions will be multiple choice and essay Questions will include topics covered throughout the semester including independent amp dependent variables hypothesis testing operational de nitions control sampling types of confounds extraneous variables internal amp external validity and what factors affect each reliability validity type of designs including surveys amp naturalistic observation within subject between subject factorials quasiexperimental ex post facto etc All of these types of questions will relate to the new information you have learned in the chapters below Thus the exam is somewhat cumulative due to the nature of research methods and designs However most of the questions will come directly from the information contained in the chapters 12 13 and 4 Chapter 12 1 Field research 0 Can increase generality and relevance of research 0 Allows us to study unplanned events 0 Able to control for fewer of the factors that in uence the behavior of our participants and thus less able to rule out alternative hypotheses o Increases external validity but decreases internal validity 2 External validity vs internal validity issues 0 External Validity 7 The generalizability of an experimental outcome 0 Internal Validity 7 Is there another way to explain the outcome besides the independent variable 3 Quasiexperimental designs all when appropriate strengths weaknesses amp be able to identify or design one 0 Time Series i Examines the effects of an event that has happened to all members of the population we are studying ii Withinsubjects retest treatment posttest l Slmple good for studying something we don t know a lot about V Weaknesses 1 Don t know the normal uctuation between any 2 measures pretest and posttest without a treatment imposed o Interrupted Time Series i Making several pretest and posttest measurements before and after an event ii Withinsubjects iii Pretests a treatment a posttests 4 iv Strengths 1 Better estimate of the normal uctuation from test to test V Weaknesses l Unclear whether factors other than the treatment caused differences in preand posttest measures why no control group 0 Multiple Time Series i Includes a control group and has multiple pretests and posttests i Betweensubjects iii Strengths 1 Rules out some alternative interpretations iv Weaknesses 1 Unknown variable may affect our experimental or our control group 0 Nonequivalent BeforeAfter Design i Difference between pretest and posttest scores are compared between treatment and control group ii Betweensubjects iii Strengths 1 Can make comparisons between 2 groups that we suspect differ before the study begins 2 Some threats to internal validity can be ruled out iv Weaknesses l Selectionmaturation still threatens internal validity o Retrospective amp EX Post Facto Designs i Work backward 7 we know the outcome and wish to determine the antecedents of this outcome ii Correlational study iii Strengths 1 Determine whether there are meaningful relationships between events that have occurred in the past 2 Popular in studies examining educational techniques disease and psychopathology 3 Good for testing alternative hypotheses iv Weaknesses 1 Because it is correlational we can only say whether a change took place not why 2 Cannot make causeeffect conclusions Correlational procedures pros amp cons when appropriate to use limitations of interpretations of results cumulative 0 Pros Efficient alternative and an important initial step in the inferential process i39 Lack of correlation can rule out possible causality o Cons Third Variable Problem 7 there could be another variable in uencing the two you are looking at 0 When to use If there is a process we think is important to understand but do not know enough about it to design a highly controlled experiment To show the relationship between to variables without attempting to show that one variable in uences the other 0 Limitations of Predictions i Cannot determine causality 5 Naturalistic observation pros amp cons how should conduct when appropriate to use limitations of interpretations of results 0 Pros i Describe a behavior in reallife settings ii How behavior unfolds over time o Cons Qualitative not quantitative i Subjective judgment of the researcher iii Limited representativeness of the sample iv Cannot get informed consent 0 How to Conduct i Need to include 1 Unobtrusive Observation 2 Accuracy of Observations 3 Good Data Recording ii Need to avoid l Boredom of Researcher 0 When it s Appropriate i When little is known about a phenomenon ii When it wouldn t be ethical to manipulate the variables 0 Limitations i Inappropriate to generalize if you don t examine other instances of the same phenomenon ii Does NOT provide info on how 1 variable in uences another 0 Interpretations of Results i Give a mass description of a phenomenon ii Gains insight about patterns of lawful relationships iii Determine whether a relationship possibly exists 6 For all designs in this section be able to determine if a design carried out in a scenario is appropriate or whether another design may be better Chapter 13 7 Types of singlesubject designs all when use how data is presented pros and cons 0 Case studies naturalistic amp oneshot i Naturalistic 1 When to Use a When you want to focus on problematic or exceptional behaviors 2 Data a Descriptive narrative by the researcher b Qualitative self report 3 Pros a Valuable descriptions of new phenomena b Leads to insight and the proposal of models 4 Cons a Cannot make strong inferences b Cannot make causal claims c Selective forgetting d Deceptions ii One Shot 1 When to Use a Observe right after a treatment 2 Data a Direct observation of subject after a treatment 3 Pros a Valuable descriptions of new phenomena b Leads to insight and the proposal of models c More experimental that naturalistic design 4 Cons a Cannot make strong inferences b Cannot make causal claims c Only get one shot Experimental singlesubject designs reversal design multiplebaseline design multielement design i Reversal Design 1 When to Use a When you want to show a different pattern of responding when a treatment is present vs not present 2 Data a Baseline treatment baseline b ABA 3 Pros a Multiple observations at each stage b Can treat groups as a single subject 4 Cons a Only works when the treatment on behaviors reduces to baseline quickly ii MultipleBaseline Design 1 When to Use a When you believe that your treatment will permanently change the behavior or be long lasting 2 Data a Monitor several behaviors of a single subject simultaneously 3 Pros a Can determine if treatment was specific to a particular behavior 4 Cons a Each behavior being monitored must be independent of the others iii MultiElement Design 1 When to Use a When you want to examine two or more levels of your IV 2 Data a Compare different levels of a given IV b Incorporate many reversals c ABAB 3 Pros a Frequent reversal of treatments i Can examine changes when background variables exist that cannot be changed Differences in terms of the effects of each treatment emerge quickly 4 Cons a Best for treatments that are shortlived b Sequence effects 8 Evaluating singlesubject experimental results 9 Identify whether design appropriate in a particular scenario Chapter 14 10 Ethical considerations in research 0 Know what can cannot due 0 Be able to identify if study is ethical or not 0 If a study violates ethics identify which ones 0 When ethics are violated know how to change a study to make it ethical ll Participants rights 0 Voluntary participation be able to identify if truly voluntary i Freedom to join the study ii Freedom to leave at any time o Informed consent know all components of i Description of the research which includes 1 The general purpose 2 The potential risks ii Inform them of any factor that might in uence willingness to participate iii Give to participant before participation 0 Right to privacy know some procedures that will insure privacy i Thoughts feelings performance of the participant should not be made public without the participant s consent ii Con dentiality 1 Cannot release data of a personal nature to anyone else without the participant s consent iii Anonymity 1 Personal identit be ke se arate from one s data l If study lowers any aspect of human functioning ii APA guidelines iii IRB 12 Scientists rights 0 Guidelines of APA American Psych Assoc see text for details 1 0 Use of deception in research when okay when not okay procedures to reduce risks i The bene ts must outweigh the risks ii Must debrief thoroughly iii Must avoid it if you can 0 Example risk studies i Milgram study ii Radiation study iii examples in lecture 0 Can seek answers to questions 13 RisldBenefit Ratio 0 Needs to be low i Bene ts high ii Risks low 14 Debriefing know all aspects 0 When to use i Provide at the end of the study 0 What content should contain i Explain the true purpose of the study ii Describe the manipulations and methods in more detail iii Can be written or verbal 1 Must be verbal if it is a deception study 0 Purpose i Important to ensure that the participant leaves the study with at least as much self esteem as when they entered the experiment ii Must leave with no anxiety iii Make sure participant does not leave in lower psychological state than when entered study iv Participant can ask questions 1 Researcher can clear up misconceptions v Researcher can ask questions 1 Strategies or factors that may have in uenced participants performance 15 Participants as colleagues o A participant should be treated as one treats a colleague as someone who is valued and respected for the information that heshe can offer 16 Animals as subjects see speci c outline in text 0 If certain experiments are not carried out then people who might have benefitted from the knowledge gained through those experiments may be forced to endure suffering or pain that could have been avoided 0 Use when i We are interested in animals for their own sake ii What we learn about animals sheds light on human evolution iii When a certain process may be highlighted or exaggerated in animals and therefore easier to notice than it is in humans iv When the underlying mechanisms of behavior are similar across species and sometimes are easier to study in a nonhuman species v When certain experiments are impossible to do with humans because of le al or ethical restrictions l7 18 Ethical amp moral uestions l9 quot 39f I 20 IRB institutional review board know role and what purpose is see details in text 0 Determine whether the participants i Welfare and rights are protected ii Will be exposed to potential risks iii Have free choice concerning participation iv Have informed consent 21 Responsibilities of Experimenter 0 Report accurate results Create a sound study Do not fabricate data Do not plagiarize Responsible for ethical treatment of subjects etc see notes AND text ALSO KNOW General information related to writing a scientific paper goal of each section of paper amp what information is contained in each section FOR A REVIEW CAN LOOK AT Chapter 15 or see laboratory notes and APA manual These are mainly a list of topics When studying please focus on anything that was discussed in lecture in great detail especially if it concerned a class activity 1 Also note that some information on the exams concern activities you are presently carrying out in the laboratory including the laboratory assignments Finally know all info in text book covered in the chapters listed above Don t Stress Study Hardll If you have any questions come see the instructor or Lab TA You can do well
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'