Methodology and Statistics - Introduction_T02 (1).pptx

Methodology and Statistics in Psychology: Introduction Topic 02 First steps in statistics – descriptive statistics and data visualization Psychology does not study „real life”…? Wrong assumption: psychological studies are useless unless they are conducted in natural environment Example: biofeedback,...

Methodology and Statistics in Psychology: Introduction Topic 02 First steps in statistics – descriptive statistics and data visualization Psychology does not study „real life”…? Wrong assumption: psychological studies are useless unless they are conducted in natural environment Example: biofeedback, memory / attention research Mass media and polls People expect that scientists (including psychologists) will conduct studies using only representative samples, same as companies that conduct polls Psychology does not study „real life”…? Two types of selecting participants I. Sample taken from the population II. Random assignment to groups Is the first type always necessary? Psychology does not study „real life”…? How about other sciences? Animal studies Rats Monkeys Mice Fruit flies …are not representative for their species! Theory-driven research vs direct application Is direct application possible? Politics vs science – fierce fight! Theory-driven research are not meant to be applied immediately Interesting coincidences: Rats – arthritis - ulcers Viagra Take aim…fire! Miss! Politicians vs scientists William Proxmire (US Senator) „Why monkeys clench their teeth?” Research on stress– indicators allowing to measure stress in people who spend a lot of time with other people in small, closed spaces But also… 2023 Przemysław Czarnek object to a NCN grant: Trans femininity and sadomasochism/BDSM. Relationships and tensions in the field of gender production (Jan Szpilka) Breakthrough vs converging evidence Another wrong expectation from the public Public expects that change in science happens because of single „great studies” Connectivity principle New theory should not only explain new data, but all of the old data as well Be careful of pseudoscience! Strategy 1: make a theory infalsifiable Strategy 2: old data are irrelevant (new theory transcends them), but there are no data to confirm the new theory (it is… welll… too new) Converging evidence There are no perfect studies Analyzing many studies shows patterns of results Sciences advances on the base of those patters, even though each study has its faults. Converging evidence Science vs… business TV, computer games and violence Talking on the phone while driving Smoking Smokers are 15 times more likely to die of lung cancer than non smokers (Stanovich, 2010) Climate change Evolution What business does: – Make an impresion that a specific pattern rest on one crucial study – Criticize this study Multiple causation Each behavior/event has many causes Interactions (coexsistence of two or more causes increases risk more, than the sum of risks for these causes) Plane crashes Depression: genetics + trauma Antisocial behavior: fenetics + child abuse, problems at home, problems during birth, etc. Multiple causality… …and wrongfully ignoring it J.R.R. Tolkien 1892 - 1973 Smoking Job market Academic success Probabilistic nature of scientific knowledge It is the same in psychotherapy – we experiment, in the sense we are not able to guarantee that a specific technique will work well for a specific patient C.S. Lewis 1898 – 1963 Field, 2017, p. 39. The language of science Theory Researcher’s beliefs about how the world works, about how or why a particular phenomenon occurs. Hypothesis Precise statement of assumed relationship between variables. Research question The question a researcher is trying to answer in an investigation. Research prediction Prediction in precise terms about how variables should be related in the analysis of data if a hypothesis is to be supported. The last two are often used interchangeably, although the distinction above is correct and useful. (EXAMPLE: heat and aggresion) The language of science Theories cannot be proven Proves => Suggests Evidence to suport or contradict the theory Findings (the what?) Conclusions (the why?) Empirical observation – investigating phenomena without preconceptions (Francis Bacon) Popper (1959): theories must be falsifiable How potentially could we falsify this theory? Coolican, 2019, p.9 Hypotheses Our hypotheses may be non-directional or directional NON-DIRECTIONAL: there are significant differences in the running speed of dogs and cats DIRECTIONAL: dogs run significantly faster than cats SPOILER ALERT: We will talk more about this in the future! Variables, variables, variables… Type/nature of the variable What is being measured and how? Qualitative: nominal and ordinal scales Quantitative: interval and ratio scales Role that a variable plays in a study: Independent variable Dependent variable Predictor variable Outcome variable Where do we start? Both options seem ok  We will go with role first Role of the variable Field, 2017, p. 46-47 Independent variable: A variable thought to be the cause of some effect. This term is usually used in experimental research to describe a variable that the experimenter has manipulated. Dependent variable: A variable thought to be affected by changes in an independent variable. You can think of this variable as an outcome. Predictor variable: A variable thought to predict an outcome variable. This term is basically another way of saying ‘independent variable’. (Although some people won’t like me saying that; I think life would be easier if we talked only about predictors and outcomes.) Outcome variable: A variable thought to change as a function of changes in a predictor variable. For the sake of an easy life this term could be synonymous with ‘dependent variable’. Independent and dependent variables Memory and Sleep Independent Variable (IV): Amount of sleep participants get. Dependent Variable (DV): Number of words participants can recall from a list they studied the previous day. Experiment Description: Participants are divided into two groups. One group sleeps for 8 hours, and the other group is sleep-deprived (only gets 4 hours of sleep). The next day, both groups are given a list of words to study and later asked to recall as many words as they can. Effects of Reward on Task Performance IV: Type of reward (monetary vs. praise). DV: Number of tasks completed correctly within a given time frame. Experiment Description: Participants are asked to complete a set of puzzles. One group is promised a monetary reward for each puzzle solved, while the other group is promised verbal praise. Independent and dependent variables Stress and Cognitive Performance IV: Exposure to a stressor (e.g., a loud noise or a difficult math problem). DV: Performance on a subsequent cognitive task. Experiment Description: Participants are divided into two groups. One group is exposed to a stressor, while the other group isn't. Both groups then perform a cognitive task, and their performance is measured. Effects of Caffeine on Alertness IV: Amount of caffeine consumed (e.g., no caffeine, 100mg, 200mg). DV: Scores on an alertness or reaction time test. Experiment Description: Participants consume drinks with varying amounts of caffeine and then perform a task that measures their alertness or reaction time. Predictor and outcome variables Studying and Academic Performance Predictor Variable: Number of hours spent studying per week. Outcome Variable: GPA (Grade Point Average) at the end of the semester. Study Description: Researchers want to determine if the amount of time students spend studying per week can predict their academic performance, as measured by GPA. Social Media Use and Mental Well-being Predictor Variable: Average hours per day spent on social media. Outcome Variable: Scores on a well-being or depression scale. Study Description: Researchers aim to determine if the amount of time individuals spend on social media platforms daily can predict their levels of well-being or symptoms of depression. Predictor and outcome variables Childhood Socioeconomic Status and Adult Health Outcomes Predictor Variable: Socioeconomic status during childhood (e.g., low, middle, high) based on factors like parental income and education. Outcome Variable: Presence of chronic health conditions in adulthood. Study Description: The goal is to determine if socioeconomic status during one's childhood years can predict health outcomes in adulthood. Physical Activity and Cognitive Function in Older Adults Predictor Variable: Number of days per week engaging in physical exercise. Outcome Variable: Scores on a cognitive function test. Study Description: Researchers investigate whether the frequency of physical exercise in older adults can predict their cognitive function. Qualitative and quantitative scales Depending on the measurement scale used, we can collect different types of information, with varying degrees of accuracy. Nominal scale Ordinal scale Interval scale Ratio scale QUALITATIVE QUANTITATIVE LET’S UNPACK IT ONE BY ONE! Qualitative: nominal Allows classifying objects into different categories... and that's it. Allows stating that objects belong to the same group or to different groups. Does not allow ordering information (e.g. ascending). Least informative, e.g.: Gender. Eye color. Car brand Nationality Control/experimental group (!) NO UNIT OF MEASUREMENT Qualitative: ordinal It allows for organizing observations in terms of the intensity of the measured trait. We cannot determine how much one observation differs from another. Young people (20-25 years old) Middle-aged people (40-45 years old) Elderly people (60-65 years old) NO UNIT OF MEASUREMENT efully easy to remember… t’s a risky example! Due to range, variance, etc. groups should not overlap! Quantitative: interval A tricky scale  It allows determining exactly how different two observations are. There is no absolute zero point (e.g. Celsius degrees). The scale has a constant unit of measurement (equal intervals between values). In the land of minor methodological abuses: subjective assessment scales… (like… are units really constant?) Psychological measurements: anxiety, mood, focus, etc. Quantitative: ratio It has an absolute zero point: zero on a ratio scale means there is a total absence of the variable you are measuring. It informs us not only "how much A differs from B", but also how many times A differs from B (e.g., John earns three times as much as Mike). Earnings Reaction time (or time in general – hours, seconds, etc.) Speed Number of points on a test Some words of wisdom… NOTE: We always try to collect data on the highest possible scale. For example, age as the number of years, not as a choice from several categories. What is the difference between an aquarium and fish soup? From a higher scale, we can descend to a lower one, but not the other way around. Descriptive statistics These are numerical characteristics of the distribution of a given variable. Their task is to summarize the collected data using numbers. Measures of central tendency describe the center position of a distribution for a data set. The most common measures include the mean (average), median (middle value), and mode (most frequent value). Measures of dispersion describe how spread out the values in a data set are. Common measures include the range, variance, and standard deviation. Measures of distribution symmetry assess whether the values of a dataset are symmetrically distributed or skewed to one side. Common measures include skewness and kurtosis. Measures of central tendency MODE Most frequent value in a data set 2 2 3 3 3 3 3 3 3 3 4 4 5 5 6 Mode = 3 There can be more than one! 2 2 2 3 3 3 3 3 4 4 4 4 4 5 5 6 CAN BE CALCULATED FOR EVERY MEASUREMENT SCALE Measures of central tendency MEDIAN (Me) Value that divides an ordered data set into two (middle value) 2 2 3 3 3 3 4 4 5 Me = 4 5 6 6 6 Sometime it has to be calculated 2 3 4 5 6 7 Me = (4+5):2 = 4.5 CAN BE CALCULATED FOR EVERY MEASUREMENT SCALE, EXCEPT THE NOMINAL SCALE… Measures of central tendency MEAN Arithmetic mean, average value (M): It is the sum of all values of a variable divided by their numer 2 3 4 5 6 M = 20:5 = 4 CAN ONLY BE CALCULATED FOR QUANTITATIVE SCALES… WHY?  Measures of dispersion RANGE The term 'range' refers to the 'distance' between the lowest and highest score on a given variable. It is useful for catching erroneous entries 2 3 3 4 5 6 7 8 RANGE = 8 – 2 = 6 8 8 Measures of dispersion VARIANCE It is a measure of the dispersion of scores around the mean Variance is equal to the sum of the squares of deviations from the mean, divided by the number of scores minus one. This guy again… Measures of dispersion 2 3 3 3 3 4 4 4 5 5 6 1. We calculate the mean 2. We calculate the distance of each result from the mean 3. We square the differences 4. We add the squared values 5. And we divide the sum of squared values by numer of participants minus 1. 13,64 / 10 = 1,364 Not so bad after all, right? Measures of dispersion STANDARD DEVIATION (SD) Square root of the variance √ 1,364 = 1,17 NOTE: Variance is expressed in squared units, standard deviation is expressed in 'normal' units. For example: zł2 (variance) versus zł (SD). And because of that SD is easier to interpret  Measures of distribution symmetry Hmm… We will deal with them later, as we first need to talk about normal distribution (Topic 3)… meaning this little bugger… stay tuned!  Mean and Standard Deviation An interesting relationship… Knowing the mean without knowing SD is rarely very useful The mean is susceptible to outliers, e.g.: Mean income in a company: the mean may be high, simply because it includes the salaries of the Board membersand employees earn much less Field, p. 65 Basics of data visualisation Carefully select the type of plot The plot/graph should be helpful, if it is not, do not use it  THIS! Respondent's gender male female Respondent's age NOT THIS! 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68 71 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 Marital status married widow divorced separated never been married Marital status 3000 2500 Frequency 2000 1500 1000 500 0 married widow divorced Marital status separated never been married