ITRIP Lecture 6 - Correlational Research Design 2024 PDF

SCIENTIFIC FOUNDATIONS OF PSYCHOLOGICAL SCIENCE Lecture 6: Correlational Research Design Dr Melanie Murphy [email protected] Reading: Howitt and Cramer: Chapter 11 – Cross sectional or Correlation Research (NonManipulation Designs Field (2017) Chapter 1 (1.7) and Chapter 8 Salkind: Chapter 9 –...

SCIENTIFIC FOUNDATIONS OF PSYCHOLOGICAL SCIENCE Lecture 6: Correlational Research Design Dr Melanie Murphy [email protected] Reading: Howitt and Cramer: Chapter 11 – Cross sectional or Correlation Research (NonManipulation Designs Field (2017) Chapter 1 (1.7) and Chapter 8 Salkind: Chapter 9 – Correlational Research DDP Week! DDP presentations are this week Please ensure you finalized and submitted the presentation before your class. You can use Zoom to practice together outside of class. If your class is online, you will need to assign yourself to one of two breakout rooms for the presentation. The tutor will instruct you how to do this Your feedback after the presentation will include review of what you presented as well as guidance on future research design decisions. During class, you will provide feedback to other groups via a form in a link shown by the tutor during class You have until Wednesday 24th April to submit your group contribution mark via a form sent by your tutor to your LTU email Students must provide marks to receive marks for these components of the task* WHAT HAVE WE COVERED IN ITRIP SO FAR? Definition of concepts is important for scientific communication. Constructs are concepts operationalized. More precise operationalisation allows for more precise measurement. Observable, quantifiable Different scales of measurement provide different amounts of information Nominal, ordinal, interval, ratio The score we see (Observed Score) includes the ‘True’ Score and Measurement Error The amount of measurement error relates to the reliability and validity of the experiment Reliability is the extent to which a score is consistent (i.e., reproducible) across time and between observers Validity is the extent to which the score is consistent with theoretical expectations about how the construct should behave Internal validity External Validity Experimental Research seeks to manipulate variables to find differences True experimental research has randomised control design the true experimental method has the most, and the quasiexperimental method is somewhere in the middle. The more control a design allows, the easier it is to attribute a cause-and-effect sequence of events. Another way in which these three designs differ from one another is the degree of randomness that enters into the design. You already know that the word random implies an equal and independent chance of being selected, but that definition and concept can be applied beyond the selection of a sample of subjects from a population to the concept’s importance in experimental design. quasi-experimental designs. Even though quasi-experimental designs will be discussed in Chapter 12, it is included here so you can see a comparison of all design types. Notice that many of these differences focus on the process of randomization of selection procedures, subjects, and assignment. WEEK 5: EXPERIMENTAL Pre-Experimental Designs DESIGN SUMMARY Pre-experimental designs are not characterized by random selection of participants from a population, nor do they include a control group. Without either of these, the power of the research to uncover the causal nature of the relationship between independent and dependent variables is greatly reduced, if not entirely eliminated. The point at which random assignment enters the process These designs allow little or no control over extraneous distinguishes different types of experimental designs from one another. Elimination (holding other variables that might be responsible for outcomes constant), or minimisation, of than what Use of a control group andthe researcher intended. For example, a parent other possible explanations of Method of investigation randomisation uses an old folk remedy (wearing garlic around Actually,the different stepsand need to be taken to ensure the of participants the effect. If this is the not neck) determines internal toexperimental treatments are to hallmarks ward offof the evil spirits associated with a child’s cold. quality ofvalidity true randomness in the best of all possible, distribute external of the study a true experiment (‘balancing’) thesetype of Lo and behold, it works! This is the weakest designs. influences between groups. experimental conclusion to reach because there is virtually The first step is one you know most about, the random no comparison to show that the garlic worked better than selection of subjects from a population to form a sample. This is Table 11.1 Differences between pre-experimental, true experimental, and quasi-experimental designs. Condition Pre-Experimental Design True Experimental Design Quasi-Experimental Design Presence of a control group? In some cases, but usually not Always Often Random selection of subjects from a population? No Yes No Random assignment of subjects to groups? No Yes No Random assignment of treatment to groups? No Yes No Degree of control over extraneous variables? None Yes Some Fig. Salkind (2018). Pg 184 CORRELATIONAL RESEARCH Description of relationships between variables NOT a cause-and-effect relationship Used in exploratory research to provide ideas about level of association to develop hypotheses that can be tested by experimental research Amenable for use in large surveys X Y Experimental vs Correlational Research Correlational Research Uses of a group sampled with varying amounts of one characteristic and another Looking at the relationship between these characteristics Experimental Research Uses a (homogenous) sample of participants who are randomly allocated to groups of treatments (X), including a control treatment, and tested on an outcome measure Manipulation of groups to study its effect on outcomes, while keeping other factors constant. Correlational Research Figure 11.1 Howitt and Cramer CORRELATION: WHAT DOES IT MEAN? Simply, that two concepts we have measured have something in common. It could be that: X causes Y or, Y causes X or, X and Y are caused by a third variable CORRELATION DOES NOT IMPLY CAUSATION! The finding of a significant relationship does not tell us anything about whether one variable influences the other. Nor does it provide information about the meaningfulness of the relationship. i.e. Is chocolate the only thing that can explain why some countries have more Nobel Prize winners? WHAT IS A RELATIONSHIP? The relationship between two variables is the manner and extent of how their scores covary. The correlation coefficient provides a single numerical estimate of the strength and direction of the relationship between two variables An index of this relationship is the correlation coefficient. The Pearson’s product-moment r is the most commonly used index. WHAT IS A RELATIONSHIP? The line of best fit gives a visual representation of this relationship. We can use it to predict a value of one measure based on the score in the other. For example, if a child is a particular age, what could we predict their height might be? Clinical example; if we knew a child’s age, what level of reading ability would we expect? Correlation Coefficient: r If X… And Y… r is Example positive The taller one gets (X), the more one weighs (Y). positive The fewer mistakes one makes (X), the faster a task is completed. negative Greater levels of wellbeing (X), the lower level of depression (Y). negative The less anxiety (X), the more confidence (Y). Adapted from Table 9.2 (Salkind, 2009) WHAT IS THE NATURE (DIRECTION) OF THESE RELATIONSHIPS? a a). b). c). b c Direction & Magnitude of r Direction + X increases, Y increases X increases, Y decreases Magnitude.8.6.4.2.0 - 1.0 -.8 -.6 -.4 -.2 Very strong Strong Moderate Weak Very weak Strong (positive) correlation Y 100 80 60 40 For example, 20 Y = overall health score X = index of happiness 0 2 4 6 8 X 10 Strong (negative) correlation For example, Y 100 Y = motor coordination X = drinks consumed 80 60 40 20 0 4 8 12 16 20 X Near perfect (positive) correlation Y 1.0 0.8 0.6 For example, 0.4 Y = Fuel efficiency X = KM driven 0.2 0 0.2 0.4 0.6 0.8 1.0 X No correlation Y 150 125 100 75 For example, 50 Y = IQ X = tea consumption 0 8 16 24 32 42 X Factors that affect correlations Non-linearity Outliers Sampling from a restricted range Heterogeneous subsamples Non-linear trends… r r Quadratic Cubic ( 1 bend) (2 bends) An outlier strengthening a correlation Y r =.70 (with outlier) r =.10 (without outlier) FB: Dandy Memes for the Dank Darwinian X From Stevens (1992, p.16) An outlier weakening a correlation Y r =.80 (without outlier) r =.20 (with outlier) X FB: Statistical Statistics Memes Sampling from a restricted range Y Sample 1 Sample 2 X Different findings will arise from the two different samples Heterogeneous subsamples Y 0 X Different findings will arise from the two different samples Shared Variance (r 2 ) Y X r2=0 X Y r =.25 2 X Y r 2 =.09 X Y r 2 =.64 Degree to which the measures have something in common Threadless.com METHODS OF COLLECTING CORRELATIONAL DATA Field observation recording behaviours during clearly specified sampling periods (onset, offset) Questionnaire surveys interviews mail internet Secondary data analysis historical records (e.g., medical records, newspapers, archival data) How representative is the sample? RANDOM SAMPLING AND SAMPLE SIZE Random sampling is used to obtain the data that accurately reflects the characteristics of the population being studied. Larger samples are more likely to produce data that is representative of the population. Simple sampling Sampling methods include: Note: Typically, except for epidemiological studies, sampling is opportunistic (e.g., volunteers). A biased sample is one in which the data is not representative of the population. Systematic sampling Stratified sampling Concept definitions Theory (Nominal definition) (Operational definition) Real World HOW RELIABLE AND VALID ARE THE MEASURES? Both measures should be: Reliable (i.e., reproducible) Valid (i.e., vary in a manner consistent with theoretical expectations about how their constructs should behave) CAUTIONS FOR INTERPRETATION FOR CORRELATION ANALYSIS The “Third Variable Problem” A lurking variable! There could be another variable that exists that influences the two variables you are looking at that you didn’t measure or take into account. Extrapolation of findings To apply the findings to other situations/samples, one must be careful that they have similar characteristics to the initial study. (Source: http://xkcd.com/552/) “Significant linear correlation (r =.79, p <.001) between chocolate consumption per capita and number of Nobel Prize winners per 10 million persons”. Messerli, F. H. (2012). Chocolate Consumption, Cognitive Function and Nobel Laureates. New England Journal of Medicine, 637 (16), 1562-1564. Impact of situations Relationship between X (e.g., work) and Y (reward) may change across situations (e.g., low income vs high income jobs). population concept X X Y situations concept Y X-Y RELATIONSHIPS IN DIFFERENT SITUATIONS X-Y relationships may be different across situations and groups of people Cultural considerations Ecological considerations Beyond simple correlation … Degree of theoretical elaboration Path Analysis Hierarchical Regression Multiple Regression Simple Regression Correlational research is the study of relationships between variables (Pearson’s r, ranges between -1 and +1) SUMMARY It cannot infer a cause-and-effect relationship unless you can present a logical argument, aside from the data, that X (e.g., rainfall) is the cause of Y (e.g., plant growth) It is used in exploratory research to provide ideas of cause-and-effect hypotheses that can be tested by experimental research WEEK 7: INTRODUCTION TO QUALITATIVE RESEARCH DESIGN REVISION MID-SEMESTER QUIZ Multiple choice quiz presented via on LMS in Week 8. 9am Friday 3rd May to 9am Saturday 4th of May (AEST) It will be open for 24 hours, and once accessed you will have 45 minutes to complete all questions*. Lecture Stream Lecture Titles/Content Philosophical Foundations of Psychology Ethical and Cultural Foundations of Psychology Number of Items How do we know what we know? Is psychology a science? How do we know what we know? How does psychology progress? 2 2 The ethics of psychological research Understanding the relationship between culture and psychology Cultural research Supported decision-making Introduction to Research in Psychology Concepts and Measurement Reliability and Validity of Measurement Experimental Research Design Correlation Research Design Internal and External Validity 2 2 2 2 2 Tutorial Questions Questions related to exercises and methodology content 3 Practice Questions Random selection from practice questions 5 2 2 2 2