Class 20- nov 9 -Sample vs population in terms of smarties 1. Population level: amount of each color of smartie made in factory 2. Probabilistic trend: expected proportion of each color of smartie in the population at factory 3. Random sample: in any packet of smarties, how many of each color there -As sample size increases, pattern in sample better represents/estimates truth in the population. Samples are imperfect assessment of overall probabilities. As sample size decreases, estimate is less accurate -Probabilistic trend: 1. What is true overall effect? 2. Not likely to be reflected in every sample and case 3. Ex: regression line as the best fit to summarize all data points -Inferential statistics: 1. Using data collected from a sample to infer what's happening in population. 2. Is effect found in sample due to random chance or a true effect in the population? -Why often wrong: 1. Focus on (random) rare events (base rate fallacy)- failure to use probabilistic info when making judgements, focus on specific information rather general (Ex: person who ... ) 2. Seeing patterns in randomness- see trends in events that are really due to chance, looking for patterns in coincidences, gambler's fallacy Class 21- nov 14 -Small samples subject to more error in estimating population value -Even with random assignment, problem still remains -Random assignment works best with large sample sizes, since chance contributes a lot to statistical analysis and research methods -When an effect does not exist in the population, but we conclude that there is an effect -> Type 1 error -North Dakota wine study: how people's perception of origin of wine affect how much they like it -DV: expectations of wine, ratings of wine, portion of companion food eaten -Generally California wine rated more positively than North Dakota
-Non directional hypotheses Null Hypothesis Mean 1 = Mean 2 Ho · True effect: Label does not affect taste perception in the population · Random chance caused any difference between groups in our sample -Directional hypotheses Null Hypothesis . Mean 1 ? Mean 2 Ho . True effect: California label will not make wine taste better in population . Random chance caused Mean 1 > Mean 2 in our sample Research Hypothesis · Mean 1 + Mean 2 · H1 or HA · True effect: Label affects taste perception in the population · Random chance is a very unlikely explanation of the difference between groups in our sample Research Hypothesis . Mean 1 > Mean 2 · H1 or HA . True effect: California label makes wine taste better in population . Random chance is a very unlikely explanation that Mean 1 > Mean 2 in our sample -Key part of interpretation: was the difference due to chance or does it reflect a real difference in the population? -Whether something occurred due to chance is the most parsimonious alternative explanation for any research finding. -When analysing data, start with assumption that null hypothesis is true. Can we reject the null hypothesis? -Statistical significance: 1.