Continuation
This exercise is a continuation of this data wrangling exercise and, thus, depends on data frames constructed there. Please load/run your script from that exercise to access the Yellow Perch captured in the 1990s data frame.
Basic Analysis I
- Construct graphs appropriate to answer the following questions.1
- Describe the relationship between weight and length (in mm here and throughout).
- Describe the relationship between log-transformed weight and length.
- From the plots above there is a clear minimum length for which the weights were precisely obtained. What is that length? [Reduce the data frame to fish greater than this minimum length for the questions below. You should also remove all fish for which a weight was not recorded.]2
- Compute the weight-length relationship with an appropriate linear regression.
- Plot the results (data and the fitted relationship) on both the transformed and raw scales.3 Comment on the fit.
- Construct a residual plot.4 Comment.
- Express your results as an equation on the transformed scale.
- Express your results as an equation on the raw scale.
- Carefully interpret the meaning of the slope of the weight-length relationship.
- Is there statistical evidence for isometric or allometric growth?
1 If you completed this graphing exercise then you created the necessary graphs there.
2 There are several outliers in this data that should be corrected or removed. For simplicity, leave them in the data for this exercise.
3 This post may be useful.
4 This post may be useful.
Basic Analysis II
- Recompute the weight-length relationship using the original length in inches. How do the slope and y-intercept from this model compare to the results from the previous question?
Extended Analysis
- Construct a plot that allows you to qualitatively assess if the weight-length (in mm here and throughout) relationship differs between the three gears.
- Fit a model that allows you to determine if there is a statistically significant difference in the weight-length relationship between the three gears.
- Construct a residual plot for this model.5 Comment.
- Is there a statistically significant difference in the weight-length relationship between the three gears? Provide evidence for your findings and be very specific with your conclusions.
- Without fitting separate regressions for the three gears express the weight-length relationships on the raw scale for all gears (i.e., write three specific equations).
- Construct a plot that illustrates your findings.6
Solution Code:
Available upon request to students not in a class. Contact fishR maintainers.