class: center, middle, inverse, title-slide .title[ # M1 IREF: Econometrics ] .subtitle[ ## An introduction ] .author[ ### Laurent Bergé ] .institute[ ### B
x
SE ] .date[ ### Fall 2023 ] --- ![](data:image/png;base64,#images/nutrition_coffee.PNG) --- ![](data:image/png;base64,#images/nutrition_vitamin.PNG) --- ![](data:image/png;base64,#images/nutrition_magic_food.PNG) --- ![](data:image/png;base64,#images/nutrition_heart_drink.PNG) --- ![](data:image/png;base64,#images/nutrition_wine.PNG) --- ![](data:image/png;base64,#images/nutrition_wine_bis.PNG) --- # Question .v-40[] .center.fs-25[Why am I showing you this?] --- # Answer ![](data:image/png;base64,#images/khoa_causality.PNG) --- # Objectives of this class - understand the principles of OLS estimation - understand what is an experimental design and why they are not all equal - be able to interpret the results of quantitative studies - understand what is causality and what it entails -- After the class, you should be able to: - use the previous knowledge to precisely decipher information - use the previous knowledge to do some good work! --- # Why study econometrics? Econometrics applications are damn important! - to answers public policy questions: - does adding a new lane reduces traffic congestion? - does class size affect learning outcomes? - to answer public health questions: - is this drug effective against covid? - what is the economic cost of national lockdowns? - to answer environmental questions: - which crops can adapt to climate change? - what crop combinations increase the yields of organic farming? -- And also on the marketing side (all big companies use econometrics): - does hiring celebrities increase sales? - does adding green labels increase sales? --- # Let's start with questions .block[{Q1} Do you think aspirin reduces, or suppresses, headaches? ] -- .block[{Q2} Do you think drinking coffee increases the risk of cancer? ] -- .block[{Q3} Which question is easier to answer? .color1[Why?] ] ??? The idea is to make them try to figure out why the answer to one problem is easier to get. This should lead to the conditions of causality - les problèmes de mesure des variables: dépendantes et indépendants - la population utilisée pour estimer les effets estimation multivariée: => see my maths notes