PSYC 301 1st Edition Lecture 6Hypothesis TestingPart 2 of the semester is about psychology: conducting experiments and interpretingresults. We will gain an understanding on how psychologists conduct their experiments andhow they get results.Outline of Last Lecture I. Random SamplingII. Z Scores & ProbabilitiesIII. One-Tailed / Two-Tailed ProbabilitiesIV. Z tableV. QuestionsOutline of Current Lecture VI. Samples and PopulationsVII. Null Hypothesis Significance Testing (NHST)VIII. QuestionsCurrent Lecture – (Professor explained this to the class by using an example of coffee & IQ)I. Samples and Populations1) Ask a question: does drinking coffee affect adults’ IQ?2) We measure variables on a sample of individualsi. Average IQ of adults who drink coffee (the sample groups have to be measuring adults who already drink coffee)3) But we want to make conclusions about populationsThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.i. All adults4) Descriptive statistics: only summarize information just about our sample5) Inferential statistics: make inferences about the population based on our sampleII. Null Hypothesis Significance Testing (NHST) – 6 step process:1) Step 1- restate the question as a research hypothesis and a null hypothesis about the populations (rearrange and rephrase the question)i. Null Hypothesis- there is no difference between groups1. Hₒ / H0 (“H-not”)ii. Research hypothesis (alternative hypothesis)- there is some difference between the groups1. HA / HA iii. The data will be consistent with only one of these hypothesesiv. Example: does drinking coffee affect adults’ IQ?1. Null hypothesis- there is no difference in IQ between coffee-drinking adults and non-coffee drinking adults2. Research hypothesis- there is some difference in IQ between coffee-drinking adults and non-coffee drinking adultsv. We ALWAYS test the Null Hypothesis because we know how to test it. We initially assume that the null hypothesis is truevi. When trying to determine if outliers belong in the current study populationvii. The central question of NHST: what is the probability of observing these (or more extreme) data, given that the null hypothesis is true?”1. If data are likely (high probability), groups are probably notdifferent (H )ₒ2. If data are unlikely (low probability), groups are probably different (HA)2) Step 2- determine the characteristics of the comparison distributioni. Comparison distribution (sampling distribution): distribution of scores for the population given that the null hypothesis is true. (we are collecting data and comparing it to this distribution)1. Follows the normal deviation; has a certain mean and SD2. We compare our sample scores to the comparison distributionii. High probability – our sample probably came from this population (H )ₒiii. Low probability- our sample probably didn’t some from this population, making our sample different from the population (HA)iv. Compare collected data against some distributionv. With inferential statistics, we are inferring the relationship betweenour sample and the population1. There is always a chance that out inference is incorrect2. Your probability will always be greater than absolute zero3) Step 3- Determine the Cutoff sample score on the comparison distribution at which the bull hypothesis should be rejectedi. Critical value (cutoff sample score)- the decision point. 1. Is a sample score exceeds the critical value, H is probably ₒfalse (your sample probably didn’t come from this population)2. If sample score is less than critical value, H probably not ₒfalse (inconclusive whether your sample come from this population; assume it dida. Coffee example: [70 > IQ > 130]ii. 2 ways to achieve step 3 that will give you the same answer:1. Choose a cut off score- critical value2. Choose a probability- alpha level (α)a. α = 0.05 is most commonb. The probability of observing data at least this extreme due to random sampling when the null hypothesis is true4) Step 4- determine your sample score on the comparison distributioni. Steps 1, 2, and 3 take place before any data is collectedii. Now that you have collected the data, it’s time to compare your sample to the comparison distribution. (Do they differ?)5) Step 5- decide whether to reject the null hypothesisi. The sample score exceeds the critical value: reject the null hypothesis and accept the research hypothesisii. The sample score does not exceed the critical value: do not reject the null hypothesisiii. Rejecting the null hypothesis does NOT “prove” the research hypothesis. The research hypothesis is supported, but it does not make it capital-T True.iv. Failing to reject the null hypothesis does NOT “prove” or “support” the null hypothesis. The results are inconclusivev. Your hypothesis is only as good as the previous data collected.vi. You can support your hypothesis, not “prove” itvii. We are not concluding anything or concluding things to be true6) Step 6- Interpret your results*i. Lastly you have to tell the reader what this means, you can just stop after rejecting/ not rejecting the null hypothesis.ii. Rejecting H : “adults who drink coffee differ in mean IQ from ₒadults who do not drink coffee.”iii. Not rejecting H : “adults who drink coffee did not differ in mean ₒIQ from adults who did not drink coffee.”iv. Be able to communicate and interpret your conclusive results.III. Questions1) How are the research and the null hypothesis different?i. In the null hypothesis it states that there is no difference, and in the research hypothesis it states that there is some difference.2) After you collect scores from your sample, to what are you comparing it?i. Comparison and population distribution3) If your sample mean does not exceed the critical value, what do you do? Why?i. Do not reject the null hypothesis; do not assume that the sample is from a different population.ii. It is not far enough away from the population meaniii. Example: “you must be this tall to ride this