Presentation is loading. Please wait.

Presentation is loading. Please wait.

Model Trendline Multi Excel 2013 V0F 1 by Milo Schield Member: International Statistical Institute US Rep: International Statistical Literacy Project Director,

Similar presentations


Presentation on theme: "Model Trendline Multi Excel 2013 V0F 1 by Milo Schield Member: International Statistical Institute US Rep: International Statistical Literacy Project Director,"— Presentation transcript:

1 Model Trendline Multi Excel 2013 V0F 1 by Milo Schield Member: International Statistical Institute US Rep: International Statistical Literacy Project Director, W. M. Keck Statistical Literacy Project Slides at: www.StatLit.org/pdf /Model-Trendline-Multi-Excel2013-6up.pdf Model Using Trendline Non-Linear in Excel 2013

2 Model Trendline Multi Excel 2013 V0F 2 Goal: Summarize association between two variables 1.Generate seven charts showing association between two quantitative variables (slides 5-11). 2.Use seven different models: linear with forecast, linear with intercept = 0, polynomial, logarithmic, power, exponential and moving average. 3.For each chart (except moving average), show trend-line, regression equation and R 2. Show title and axis for all 4.Review comparison of model R-squared on slide 12. No description of association (trend) is necessary. For details on using Trendline to build a model, see > www.StatLit.org/pdf/ Model-Trendline-Linear-Excel2013-6up.pdf

3 Model Trendline Multi Excel 2013 V0F 3 Using Chart Trendline Create Graph. Look for + Sign. Select Chart Elements. Check Trendline box. Select More Options. Select Algebraic model Check Equation & R-square [Check Forecast or Intercept]

4 Model Trendline Multi Excel 2013 V0F 4 Algebraic Models 1) Linear: Y=a+bx. Straight line, simplest 2) Polynomial: Y= a+bx+cx 2. Multiple curves 3) Logarithmic: Y=aLn(x)+b. Ratio scale. Equal ratios have equal differences Log 10 (1) = 0; Log 10 (10) = 1; Log 10 (100) = 2 4) Power model: Y=ax b [Between log & exp.] 5) Exponential: Y=ae x/b. Constant rate of change 6) Moving average: For time series

5 Model Trendline Multi Excel 2013 V0F 5 1)Linear Model w Forecast

6 Model Trendline Multi Excel 2013 V0F 6 1)Linear Model: Intercept = 0

7 Model Trendline Multi Excel 2013 V0F 7 2) Polynomial Model

8 Model Trendline Multi Excel 2013 V0F 8 3) Logarithmic Model

9 Model Trendline Multi Excel 2013 V0F 9 4) Power Model

10 Model Trendline Multi Excel 2013 V0F 10 5) Exponential Model

11 Model Trendline Multi Excel 2013 V0F 11,

12 Model Trendline Multi Excel 2013 V0F 12 Comparison of R-squared Percentage of Weight “explained by” Height 61.6% Linear model 40.6% Linear (intercept = 0) << Worst fit 62.1% Polynomial model 61.2% Logarithmic model 63.1% Power model 63.3% Exponential model << Best fit! No equation/fit for Moving-Average model.


Download ppt "Model Trendline Multi Excel 2013 V0F 1 by Milo Schield Member: International Statistical Institute US Rep: International Statistical Literacy Project Director,"

Similar presentations


Ads by Google