Topic6_Inferential_Stat_Part1_2024_1.pdf

Topic 6: Inferential Statistics & Trend Analysis Part I Dr. Krisztina Soreg   →  The ones that we include in the model to explain or predict changes in the dependent...

Topic 6: Inferential Statistics & Trend Analysis Part I Dr. Krisztina Soreg   →  The ones that we include in the model to explain or predict changes in the dependent variable Independent It is changed, controlled in the experiment → variable (X) not affected by other variables Synonyms: predictors, factors, treatment variables, explanatory variables, input variables What you want to use the model to explain or predict → the values of it depend on other variables Dependent As the experimenter changes the independent variable (Y) variable, the effect on the dependent variable is observed and recorded Synonyms: outcome or response variable Make a list of some possible independent variables which can contribute to the following outcomes: Dependent variables: 1. High rate of poverty in a country 2. Low unemployment rate 3. Very high inflation rate 4. High level of corruption in the public sector 5. Frequent usage of chatGPT in school assessments  Independent Variable: Time spent sleeping before the exam Dependent Variable: Test Score Independent Variable: Consumption of fast food Dependent Variable: Blood Pressure Independent Variable: the amount of caffeine consumed Dependent Variable: Sleep Situation: new research is released that suggests companies can keep their employees Which one is which? longer if they offer shorter work weeks. Observation: companies with shorter work weeks The rate of employees quitting have less turnover in their staffing, while depends on the length of the work companies that require more hours have higher week → dependent variable rates of employees quitting. The length of the work week, which is Two variables: rate of employees quitting & something that companies can adjust length of work week → independent variable Situation: sociologists want to know how the minimum wage can affect rates of non-violent Which one is which? crime Research method: studying rates of crime in The minimum wage is the variable areas with different minimum wages and that is affecting crime rates → comparing the crime rates to previous years when independent variable the minimum wage was lower. Crime rate is being directly affected Two variables: minimum wage & crime rate by the minimum wage → dependent variable  Association does not imply causation! Causation can only be inferred from a randomized experiment Gross Operative Profit depends on the number of rooms, sq meters, f&b, sq meters SPA, yield management rate, etc. If there is a relationship between any two variables, it may be possible to predict the value of one of the variables from the value of the other. What do we mean if two variables have a strong statistical relationship with one another? They move together They appear to be related There might be a correlation between them Remember: sometimes, however, statistical relationships exist even though a change in one variable is not caused by a change in the other.  What can we read from the graph? Observation: ice cream consumption leads to murder…  What can we read from the graph? Observation: a pirate shortage caused global warming…  What can we read from the graph? Observation: Mexican lemon imports prevent highway deaths…  What can we read from the graph? Observation: the age of Miss America is strongly correlating with the number of murders by hot objects… Rule: correlation does not imply causation!!! Just because two trends seem to fluctuate in tandem, that doesn’t prove that they are meaningfully related to one another! How is it possible? 1. It may be the result of random chance, where the variables appear to be related, but there is no true underlying relationship. 2. There may be a third, lurking variable that that makes the relationship appear stronger (or weaker) than it actually is. What should we do? With well-designed empirical research, we can establish causation! → randomization, controlled experiments and predictive models with multiple variables Thank you for your attention! [email protected]