CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. UNIT 4 Scatterplots and Correlation CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. Fri, Nov 13 Sit in your seat Everyone needs a shoulder partner! Open to p. 259 Pull out your notebook Grab a half sheet CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. Scatterplot of Music Ranking Rank the music in order from your favorite (#1) to least favorite (#8). Copy your shoulder partner’s rank onto your paper, too. Create a scatterplot of your data on the front! Write your names big on the front. Make BIG points and write names BIG at bottom!! Answer p. 259 #1c-g as a pair (you’ll have to walk around and look at other scatterplots) CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. Interpreting direction of a Scatterplot Association is similar to correlation!!! Negative correlation: pattern that moves from the upper left to the lower right Positive correlation: pattern that moves from the lower left to the upper right Correlation Coefficient Use the variable r r is between -1 and 1 If the correlation is close to a positive line, r will be close to 1 If the correlation is close to a negative line, r will be close to -1 If the correlation is nothing near a line, r will be close to 0! What to say when the r value is … CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong or “weak”. Describe the following scatterplots and estimate the rvalues: CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. Line them up Let’s make one big continuum on the board (by taping the scatterplots up here) from strong negative to strong positive. Line ‘em up! CO: Students will order scatterplots based on the strength of correlation. LO: Students will describe a correlation as being “strong” or “weak”. How did we do? Answer question p. 262 #4 with your shoulder partner about the scatterplots on the board. HW: p. 262 #5, p. 275 #11 CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Monday, Nov. 16 Groups: p. 264 #1-3 Check-in’s: 1c 1d 2b 2d 3b CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Interpreting Scatterplots 4 main things you need to talk about: Overall pattern : Direction (positive or negative association) Form (linear or non-linear- curved) Strength (how closely the points follow a clear form) Striking deviations: Outlier (outside the overall pattern) ???Clusters, Vary in strength, etc. CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Linear vs. Non-Linear Cloud/ Oval => “Linear” Curved => “Non-Linear” “Vary in Strength” … say that! CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Outliers Types of outliers Outliers for the x variable only (Point C) Outliers for the y variable only (Point B) Outliers for both the x and y variables (Point E) Outliers that are only “outliers” when you look at x and y at the same time (Points A and D) CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Correlated or Caused Correlation does not imply causation! A correlation between two variables does not necessarily mean one causes change in the other. Shark attacks and ice cream sales… On a beach in Southern California the number of shark attacks was strongly correlated with ice cream sales at a little beach shack. Does this mean ice cream is CAUSING shark attacks? What else could be going on here? CO: Students will estimate the correlation coefficient of scatterplots and find residuals using a calculator. LO: Students will state whether a linear equation will be the best model for a scatterplot in pairs. Homework p. 268 CYU, p. 276 #12-13 CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Tuesday, Nov 17 Everyone needs a calculator and a book Pull out some paper for a few quick notes CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Scatterplot on Calculator Add the data into lists Height (x) Shoe Size (y) Stat 50 5.5 64 9 70 12.5 62 7 Enter x into L1 and y into L2 Go to your stat plots 2nd y= Make sure it’s turned on Select the scatterplot Set your window Hit Graph CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Least-Squares on Calculator (Line of Best Fit) 1. Add the data into lists a) x into L1 and y into L2 2. STAT, CALC, #8 (LinReg a+bx) 3. Add this to put it in y=: VARS, , Y-VARS, FUNCTION, Y1 1. 2. Then the regression line is entered into your y= and the values for a and b are given to you. (When you write it down, round to 2 decimal places) If you turn on the scatterplot, you can see the scatterplot and the line of regression. You will have to set up a window that makes sense!!! Height (x) Shoe Size (y) 50 5.5 64 9 70 12.5 62 7 CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Interpreting the regression line The regression line: 𝑦 =.33x −11.61 “Interpret the slope and y-intercept” Height (x) Shoe Size (y) 50 5.5 64 9 70 12.5 62 7 The slope of .33 tells us that for every .33 increase in shoe size someone will get 1 inch taller. The y-intercept is what shoe size will equal when height is 0. Do they make sense? Why/why not? CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Turn to p. 282 Groups: #1-3 Check-in’s: 1a 1c 1d 1e 2b 3b (Hint, i. wants you to use your GRAPH of the line, ii. wants you to use your EQUATION of the line!) CO: Students will find a linear regression line, calculate error in predictions and find residuals using a calculator. LO: Students will state the difference between an error in prediction and a residual in pairs. Error in prediction vs. Residual Using the regression equation : Error in prediction/Residual = observed value – predicted value Error in prediction is used when the particular type of car was not in the original data set. Residual is used when the particular type of car was in the original data set. CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Error in prediction vs. Residual Which point has the highest residual? Which point has the highest error in prediction? Where are the residuals positive? Above the line Where are they negative? Trick question! None of the points! Below the line Where are the residuals 0? ON the line CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Block, Nov. 18-19 1. 2. 3. Warm-up: How can you find a residual from a scatterplot? How can you find a residual from an equation? What’s the difference between an error of prediction and a residual? CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Now for a little help Interpreting Whenever it asks you to “interpret something in context,” think about the term is using math, and then substitute in the variables of the problem! CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. What does the residual mean? What is the equation for finding a residual? Residual = Observed – Predicted The residual shows how much higher or lower a value is than was predicted. Sample sentence: “The residual shows that _______ is ________ y-variable residual number higher/lower than what was predicted.” positive / negative (above line) / (below line) CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Example Assume for this table that we find a residual of 4.23: Years Number of Birds … … … … … … What does the residual mean in context? “The residual shows that the number of birds is 4.23 higher than what was predicted.” What does the slope mean? What does the slope mean “in context” or “in terms of the problem.” What is the equation for finding a slope? Slope = 𝑟𝑖𝑠𝑒 𝑟𝑢𝑛 The slope shows how much the y- value changes when the x- variable is increased by 1. Sample sentence: “The slope shows that for every ______ that the denominator ___ increases, the ____ increases/decreases by ___. X-variable y-variable numerator CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Example Assume for this table that we find a slope of 27: Years Number of Birds … … … … … … What does the slope mean in context? The slope shows that for every 1 that the years increase by, the number of birds increases by 27. The slope shows that for every 1 year that passes by, there are 27 more birds. CO: Students will practice interpreting slopes and residuals in context. LO: Students will interpret slopes and residuals in context using fill-in-the blanks. Practice p. 285 CYU (a-c) 4.2 WS CO: Students will create scatterplots and regression lines on their calculators. LO: Students will interpret slopes and residuals in context using fill-in-the-blanks. Monday, Nov. 30 Warm Up 1. Explain what the slope means in context, given the Ounces of Water slope of 2.33. Minutes 2. If the residual is -0.846, explain what that means. (Hint: You need the sentence starters from before break!) …….. ……… CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Warm Up: Tuesday, Dec. 1 The regression line: 𝑦 =.33x −11.61 Height (x) Shoe Size (y) 50 5.5 64 9 70 12.5 62 7 Given the above information: a) Predict the shoe size of a person whose height is 55 inches. b) Find the residual for x = 64. CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Centroid The centroid is the point (𝑥, 𝑦) is the mean of all the x’s 𝑦 is the mean of all the y’s 𝑥 The linear regression line crosses through the centroid point! CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Centroid 10 11 13 15 16 31 34 37 38 45 Find the linear regression line of this data 𝑦 = 1.92𝑥 + 12 Find 𝑥 (the mean of all the x’s) 𝑥 =13 Find 𝑦 (the mean of all the y’s) Centroid: (13, 37) 𝑦 = 37 Check to see if the line goes through the centroid by plugging it in for x and y and seeing if it is equal! 37 = 1.92 ∙ 13 + 12 37 = 36.96 CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Influential Point: p. 287 Read the bottom paragraph Pairs: #3 CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Think/Pair Share p. 290 Summarize the Math CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Influential Point Go over how to delete a point out of the list, and recalculate the correlation If the correlation doesn’t change, is it influential? CO: Students will determine if a point is influential. LO: Students will describe the effect that an influential point has on the correlation. Homework p. 307 #3, 5 CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Block Day, Dec. 2 and 3 Warm Up: p. 305 #1 Friday is your EO #5 assessment!!! CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Correlation Coefficient Which has a stronger correlation? TRICK QUESTION! They are the same graph… With different scales on the axes! Our eyes are NOT very good judges of how strong a linear relationship is! So let’s learn how to calculate r! CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Calculating Correlation You can calculate r at the same time as you do a linear regression line if you do this step first: 1) Go to catalogue (2nd 0) 2) Go to DiagnosticOn 3) Press enter to turn it on 4) Now proceed to calculate LinReg CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Now it’s Your Turn! Mins of HW Quiz Scores 38 41 56 63 59 70 64 72 74 84 45 50 Find on your Calculator: 1) Scatterplot (Window: 0 – 80, 0 – 100) 1) Linear Regression Line 2) Correlation (r) CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Now it’s Your Turn! Mins of HW Quiz Scores 38 41 56 63 59 70 64 72 74 84 45 50 Find on your Calculator: Scatterplot (Window: 0 – 80, 0 – 100) Linear Regression Line y = 1.2x – 3.84 Correlation (r) r = 0.995 CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Scatterplot (15, 85) I’m adding two points. Are either of them influential? Data r Original .995 with(5,2.5) (5, 2.5) with(15,85) Add (5, 2.5) to your lists and re-calculate r. Then delete (5, 2.5) and add (15, 85) to re-calculate r See which one influences the data the most! CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Scatterplot Data r Original .995 with(5,2.5) .999 with(15,85) .129 (15, 85) (5, 2.5) So you can call (5 , 2.5) an OUTLIER, but it is NOT an “influential point.” (15, 85) is an “influential point.” What does it mean to be an influential point? CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Definition of INFLUENTIAL POINT An influential point is a point that strongly influences the value of r. That is, if the point is removed from the data set, the value of r changes quite a bit. It is not influential if it’s far away from the other points. It’s influential, if it is far away from where the linear regression would lie. CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Homework p. 298 CYU a-d, p. 310 #6a-d p. 305 #2, p. 316 #18 Friday is your Assessment on EO 5!!! CO: SWBAT calculate the correlation on their calculators. LO: SWBAT list the steps for finding correlation saying “first, then, last.” Mon, Dec. 1 Welcome Back! Hope you had an awesome break! Warm-up: p. 294 #4 and 6b