Title | SP300 Psychology Statistics and Practical |
---|---|
Author | Diva Wong |
Course | Psychology Statistics and Practical |
Institution | University of Kent |
Pages | 94 |
File Size | 1.3 MB |
File Type | |
Total Downloads | 11 |
Total Views | 737 |
SP300 Psychology Statistics and Practical Diva Wong SP300 Psychology Statistics and Practical SP300 Psychology Statistics and Practical............................................................................................................................. 1 Week 1 Method, Demo &...
SP300 Psychology Statistics and Practical
Diva Wong
SP300 Psychology Statistics and Practical
SP300 Psychology Statistics and Practical............................................................................................................................. 1 Week 1 Method, Demo & Project: Scientific Method and Logic .................................................................................... 3 Week 1 Statistics: Why stats & Intro to Variables .......................................................................................................... 5 Week 2 Reliability and Validity ....................................................................................................................................... 6 Week 2 Describing Variables .......................................................................................................................................... 8 Week 3 The Hidden Profile ........................................................................................................................................... 10 Week 3 Describing variables numerically, averages ..................................................................................................... 12 Week 4 Research Design 1 ............................................................................................................................................ 14 Week 4 Shapes of Distribution ..................................................................................................................................... 16 Week 6 Demo 2 ............................................................................................................................................................ 18 Week 6 Standard deviations and Z scores .................................................................................................................... 20 Week 7 Research Design 2 ............................................................................................................................................ 22 Week 7 Relationships between variables ..................................................................................................................... 24 Week 8 Scientific writing & APA style ........................................................................................................................... 26 Week 8 Correlation Coefficients ................................................................................................................................... 29 Week 9 Project A .......................................................................................................................................................... 32 Week 9 Samples and Populations ................................................................................................................................. 34 Week 10 Project A Discussion & Writing ...................................................................................................................... 37 Week 10 Introduction to statistical tests ...................................................................................................................... 39 Week 11 Statistical Significance .................................................................................................................................... 42 Week 11 Research design 3 .......................................................................................................................................... 45 Week 12 Calculating significance for correlation coefficients ...................................................................................... 47 Week 12 Research Design 4 .......................................................................................................................................... 49 Week 13 Research Ethics .............................................................................................................................................. 53 Week 13 Related t-test ................................................................................................................................................. 55 Week 14 Project B: Background Lecture ...................................................................................................................... 57 Week 14 Unrelated t-test ............................................................................................................................................. 60 Week 15 Project B: Writing Lecture .............................................................................................................................. 62 Week 15 Chi Square & McNemar Test.......................................................................................................................... 66 Week 16 Ranking Tests (non-parametric tests) ............................................................................................................ 68 Week 16 Single-case designs ........................................................................................................................................ 70 Week 18 Regression ..................................................................................................................................................... 74 Week 18 Demo 3 Background lecture .......................................................................................................................... 76 Week 20 Related ANOVA .............................................................................................................................................. 77
1
SP300 Psychology Statistics and Practical
Diva Wong
Week 20 Deviant Leaders and Leadership Potential .................................................................................................... 79 Week 21 Project C Decision making and social preferences ........................................................................................ 80 Week 21 Two-Way ANOVA ........................................................................................................................................... 83 Week 22 Multiple Comparisons .................................................................................................................................... 85 Week 22 Project C: Discussion and Writing .................................................................................................................. 87 Week 23 Designing and Reporting Research ................................................................................................................ 90 Week 24 Final Week Quiz ............................................................................................................................................. 93
2
SP300 Psychology Statistics and Practical
Diva Wong
Week 1 Method, Demo & Project: Scientific Method and Logic •
•
•
•
•
•
• •
•
Why we need psychological science • Interaction between mind and behaviour • How the brain interacts • Why do we need to approach it from the scientific manner rather than theories? 4 main aims • Describe: mental processes occur, semantic details of it (generate mental database, gather data) • Explain: knowing the causes of the mental state, origins (what causes depression, imbalance in neurotransmitter(how)) • Predict: predict changes in the way we behave if we understand the root of the behaviour (make prediction of low levels of certain neurotransmitter will increase risk of depression) • Control: gain the ability to manipulate and control it (know adding neurotransmitter could decreases depression) • *Example: Depression What kind of questions can science address • Ask the right questions as a lot of things are very complex • Evidence base approach ▪ Take all information ▪ Aggregate it ▪ Draw conclusion ▪ Build up information Hypothetico-deductive method • Observation: we see something interesting • Questions: additional research, look in research regarding topic • Hypothesis: formulate a statement of point(statement), a testable statement about reality, has to be able to test empirically • Prediction: Guess about reality (prediction supports the hypothesis) logical statement about the future • Experiment: Gather data • Results: Analyse data that supports hypothesis, or doesn’t support hypothesis: go back to step 2 Generalisation and Falsification • Universal generalisation can only be disconfirmed • Only takes one data point to reject hypothesis (disconfirm) • Design research that refutes our claim • If we postulate a relationship between two variables and we fail to find evidence against the relationship we would be able to confirm then the relationship actually exists Rejecting hypotheses in psychological science • Try and reject that • Null hypothesis significant testing • IN practices use statically methods to reject null hypotheses • Always some chance we are wrong, results from single study not conclusive Scientific vs non-scientific thinking • Can this be proven wrong(Falsifiable) and is it practically testable Additional criteria for scientific hypotheses • Precise ▪ Clearly defined and should not be open for interpretation • Rational ▪ Logically fits with existing knowledge • Parsimonious ▪ Choose explanation with fewest assumptions, most simple assumption How does scientific research work? • Science evaluates claims in an empirical, objective, semantic and controlled manner
3
SP300 Psychology Statistics and Practical
•
Diva Wong
• Empirical: Claim, learned by observation • Collecting evidence: Study, only one thing is manipulated • Objective: objectively quantify it • Draw conclusion justified by objective Good science: • Empirical: Learned by observation • Objective: Not dependant on personal beliefs or options • Systematic: Based on careful, step by step research • Controlled: zoning in things we are measuring and exert as much control as possible
4
SP300 Psychology Statistics and Practical
Diva Wong
Week 1 Statistics: Why stats & Intro to Variables • • • •
• • • • • • • •
Statistics is a life skill Able to apply to real world situations Statistics help us interpret and understand our data Types of Data o Observations are recorded and become our data o Data can be qualitative and quantitative o SPSS is for quantitative data Variables: Any concept that can be measured and can vary Constant: quantity that doesn’t vary Independent variable: Variables you manipulate, don’t change them Dependent variable: The variables you measure How the independent variable impact on the dependent variable There are cases you can’t manipulate a variable and, in these situation,, we use the terms predictor and outcome variable. Variables can be measured several times A case can be a person, industry, organisation or object
5
SP300 Psychology Statistics and Practical
Diva Wong
Week 2 Reliability and Validity • • • • • •
•
•
•
•
Extraneous Variable: additional factor emerging from research that isn’t accounted for, e.g. height, weight, intelligence Variation in test environment: Noise, Lighting, Smell Some extraneous variables may also be confounding variables Confounding variables: extraneous variables that alters our patterns of results Not all extraneous variables are confounding variables but all confounding variables are extraneous variables Sources of extraneous variability o Participant o Researcher o Context o Measurements Error variability o Measurements reflect a combination of constructs we are interested in, plus unexplained sources of variability o Person’s score on the measured variable = Hypothetical construct being measured + Error o Aims to minimise error scores and maximize measurement construct Error control in research o Eliminate: Control as much of the setting as possible o Keep things constant o Balance o Randomise Demand Characteristics o Extraneous cues that bias participants behaviour base d on their interpretation of the aims of the study ▪ Good/bad participant role ▪ Reactivity/ Hawthorne effect, belief of being observed change the way you behave ▪ Social desirability, show behaviour that is socially expected of them to avoid judgement ▪ Experimenter expectancies o Control DC: ▪ Single • Participant doesn’t know which condition they are in ▪ Double-blind procedure • Participant and experimenter both doesn’t know which condition the participants are participating in ▪ Unobtrusive measure and deception • Try and reduce issue of Hawthorne effect • Feeling of participant being observed ▪ Concealing the experiment • Participant do not know they are in an experiment • Prevent demand characteristics Reliability o Extent to which our measure produces consistent and stable results over time o Depends on the degree of measurement error o Try and design experiments that minimise errors o Test-retest reliability ▪ Repeat experiment and would find similar results o Split half reliability ▪ Test retest reliability but in the same session ▪ Cut data into two ▪ If both results are similar then the test is reliable o Internal consistency of questionnaires
6
SP300 Psychology Statistics and Practical •
•
Diva Wong
▪ Consistent of questions show similar results Validity o The extent a measure or observation indicates what is intended o Whether measurements reflect variable, constructs and relationships they are meant to reflect o Intangible hypothetical construct -> component variables -> Operationalisation ▪ Measurement validity • Content validity o Degree to which measure includes all aspects of a hypothetical construct • Construct validity o Degree to which measure actually and only reflects the hypothetical construct ▪ Internal Validity • Degree to which the relationship observe in our sample exits actually and only between the variables of interest • Degree to which the relationship exists purely because the independent variable causes a change in the dependent variable • Higher internal validity is the more confident is the independent variable is eliciting an effect on the dependent variable • Are we doing what we say we are doing ▪ External Validity • Degree to which the observed relationship generalises to other participants in other situations • Ecological validity: procedures of experiments are similar to real life events • Temporal validity: results apply across time • Higher external validity, the more confident is that our results can be generalised o Increasing one often decreases the other What exactly are we measuring o Intangible hypothetical construct to component variable to operationalisation o Quantifying something unquantifiable
7
SP300 Psychology Statistics and Practical
Diva Wong
Week 2 Describing Variables •
•
•
• •
Types of variable o IV = variables you manipulate o DV = variables you measure o Get idea of the effects of the IV on the DV Some time you have to look at things that you can’t manipulate o Can have ideas about the direction of the relationship between two variables o Predictor and outcome variable o Predictor variable is like IV but you can’t manipulate it due to ethical or social reasons o outcome variable is like a DV Levels of measurement o Nominal Variable ▪ Name ▪ Categorical information ▪ Using numbers but only tell us category membership ▪ Labels, don’t tell us anything about the content of the category o Ordinal Variable ▪ Use a scale that tells us something about the order of the participants ▪ It doesn’t tell us how much rank 1 and 2 differ ▪ Simply tell us the order o Interval and ratio variable ▪ Continuous measures ▪ We know the difference between the measurement points as there is equal distances ▪ Interval scales: have no absolute zero point, zero is arbitrary, no meaningful zero ▪ Ratio scales: have an absolute zero point Summary o Anything that varies can be used as a variable in research Research Methods o Between participants design ▪ Comparative design ▪ Compares the score of one group of participants with the scores of another ▪ Measures two or more separate group ▪ Need to recruit twice as many people for experiment o Within participants design ▪ Repeated measures design ▪ Researcher measures two or more variables at the same time using the same participants ▪ Measures the same group repeatedly ▪ Participants are tested multiple times ▪ Don’t need to recruit twice as many people o Experiments ▪ In a true experiment, researcher would systematically manipulate the IV ▪ Any change in outcome would be due to the change in IV ▪ Used to determine cause and effect ▪ Need random allocation ▪ Key goal: To control for confounding effects • Use random allocation to groups o Correlational ▪ Survey method ▪ Rather than manipulating variables, researchers are gathering information on different variables and observe the relationship between variables ▪ Measurement without manipulation ▪ Correlation does not equal causation
8
SP300 Psychology Statistics and Practical
Diva Wong
▪ Only can say there is a relationship Summary ▪ Within participant design that measures the same group repeatedly ▪ Between participants design measures two or more separate groups ▪ Experimental methods involve systematic manipulation of the IV ▪ Correlational methods look for relationships between variables Tables and graphs o Just want to present the key characteristic, the trends of the data o Should be simple and effective which can be done graphically o Nominal variables ▪ Display the number of cases that fall into each category ▪ Frequency tables • Different groups • Number of people in categories • Percentage • Total ▪ Pie charts • Shows the relative sides of each group based on percentages • Whole of pie shows 100% with segments of pie • Easily see what the categories are and the correspondent ▪ Bar charts • Shows the relative size of each group based on the height of the bars • Bars don’t touch as it is not continuous o Continuous variables ▪ Data which falls on a continuum ▪ Histogram • Collapsing variables into grouping rather than individual ages • Bars touch because it is continuous Key concepts o Nominal variables: Those which can be placed into group o Continuous variables: those which are given numerical values o Raw data: what you take from every participant o Descriptive statistics: what you do to the raw data to make it easier to understand o
•
•
9
SP300 Psychology Statistics and Practical
Diva Wong
Week 3 The Hidden Profile • • •
• • • • • • • •
•
•
•
•
•
Research idea about group decision making Groups may not make better decisions than individuals Hidden Profile problem o When discussing decisions regarding different alternatives, Groups often aim to reach a consensus rather than a right answer Groups tend only to discuss information shared amongst members Not all group members will have access to the same information – hidden profile Everyone has slightly different information which impacts the group decision Vital information can be missed and hidden profiles are 8 times less likely to make correct decision Understand what causes effect and overcome it Making a decision in a group that changes the decision Give more weight for own decision and ignores information against candidate of choosing Why groups fail to share information o Negotiation focus ▪ Aim to research consensus rather than right answer o Discussion Bias ▪ Shared information more likely to be discussed and repeated ▪ “Yes, I thought that too” ▪ probability talking about shared information is higher than not shared inf...