AP Stats question. HELP! I feel stupid. :(

College Confidential Community
1

<p>I usually do well in Stats but that's not the point. I'm baffled here. First person to get these questions gets my undying love.</p>

<p>The productivity of American agriculture has rapidly grown due to improved technology. Here are the data on the output per hour of labor on US farms. The variable in an index number giving productivity as a percent of the 1967 level.</p>

<p>This data is year and the year's productivity. Sorry it's squished together.
1940 21 1965 91
1945 27 1970 113
1950 35 1975 137
1955 47 1980 166
1960 67 1985 217</p>

<p>Calculate the logarithms of the y-values and extend the table above to show the transformed data.</p>

<p>Plot the transformed data on a grid. Label axes completely.</p>

<p>You want to construct a model to predict productivity in the near future. Since the previous plot shows two distinct linear pattersn, you decide to perform linear regression on a subset of the transformed data. Clearly identify which data points you will use, carry out your plan and write your LSRL equation.</p>

<p>Now transform your linear equation back to obtain a model for the original productivity data. It should be in the form of y = k * 10^bx. Write the equation for this model.</p>

<p>Compare your model's prediction for 1985 to the observed productivity.</p>

<p>FYI, CASH PRIZE IN STORE FOR THE FIRST ONE TO GET IT. NOW DO YA GET HOW DESPERATE I AM?!</p>

2

<p>To transform the data, you enter the year vs. log(productivity) in your calculator tables. But for the years, you want to simplify it, so 1940 would be 0, and then 1945 would be 5 and so on. That would give you the equation of the transformed data. Yeah, then add an extra column to show the transformed points from each year which you get from your calculator when you log the productivity values.</p>

<p>Use the scatterplot function on the calculator along with the y= to graph the data.</p>

<p>I guess for the third one you just pick some points and make an LSRL out of it (with the transformed data). So you then you do the 1-variable-stats and you write down the line they give you. I think that's what you do.</p>

<p>You have your line as log y=......... So make make that 10 to the (all the variables) power. So it would be 10 to the (log y) power = 10^(constant) + 10^ (slope). Then you just solve that to make the new line. For one thing 10^(log y) just gives you y, and then equals....</p>

<p>Yeah, that's like y - y hat kind of stuff. So you just plug in 45 for x in the equation to get an observed value. Then you compare those two values like which is lower and stuff and why. Or maybe you just have to calculate it and then do the y - y hat to get the residual point.</p>

3

<p>Uh-oh. My calculator just died...</p>

4

<p>Somebody just give me the exact answers and make my life easier. Just this once, please. <em>heart palpitates</em></p>

<p>Three more days till spring break, three more...:)</p>

5

<p>Stats is better than f***</p>

6

<p>Mmmmkay, still would be happy for an answer.</p>

7

<p>Do the work yourself.</p>

8

<p>I have one question regarding stats. I know you can use the fuction "Invnorm(" to find the Z-score if u know the area. But is there a way to do it backwards? In other words, can I find the area given the Z-score with a calculator if I don't have the chart?</p>

9

<p>normalcdf(lower, upper, mean, std dev)</p>