# predict function in regression analysis

3 messages
Open this post in threaded view
|

## predict function in regression analysis

 Hi all,   I have the following data in which there is one factor lot with six levels and one continuous convariate time. I want to fit an Ancova model with common slope and different intercept. So the six lots will have seperate paralell regression lines.I wanted to find the upper 95% confidence limit for the mean of the each of the regression lines. It doesnot seem straightforward to achieve this using predict function. Can anyone give some suggestions?   Here is my data. I only show the first 3 lots. Also I show the model I used in the end. Thanks very much!      Hanna       y lot time  [1,] 4.5   1    0  [2,] 4.5   1    3  [3,] 4.7   1    6  [4,] 6.7   1    9  [5,] 6.0   1   12  [6,] 4.4   1   15  [7,] 4.1   1   18  [8,] 5.3   1   24  [9,] 4.0   2    0 [10,] 4.2   2    3 [11,] 4.1   2    6 [12,] 6.4   2    9 [13,] 5.5   2   12 [14,] 3.5   2   15 [15,] 4.6   2   18 [16,] 4.1   2   24 [17,] 4.6   3    0 [18,] 5.0   3    3 [19,] 6.2   3    6 [20,] 5.9   3    9 [21,] 3.9   3   12 [22,] 5.3   3   15 [23,] 6.9   3   18 [24,] 5.7   3   24 > mod <- lm(y ~ lot+time) > summary(mod) Call: lm(formula = y ~ lot + time) Residuals:     Min      1Q  Median      3Q     Max -1.5666 -0.3344 -0.1343  0.4479  1.8985 Coefficients:             Estimate Std. Error t value Pr(>|t|) (Intercept)  4.74373    0.36617  12.955 2.84e-14 *** lot2        -0.47500    0.41129  -1.155   0.2567 lot3         0.41250    0.41129   1.003   0.3234 lot4         0.96109    0.47943   2.005   0.0535 . lot5         0.98109    0.47943   2.046   0.0490 * lot6        -0.09891    0.47943  -0.206   0.8379 time         0.02586    0.02046   1.264   0.2153 ---         [[alternative HTML version deleted]] ______________________________________________ [hidden email] mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-helpPLEASE do read the posting guide http://www.R-project.org/posting-guide.htmland provide commented, minimal, self-contained, reproducible code.