Question 7

The RichnessIndex dataset contains records of bird sightings observed by owners of the 'Koobabbie' property in the Wheatbelt over a 30 year period. The dataset can be used to develop measures of species diversity. The data was collected using a simple binary system, and a tick was placed indicating that a bird species was sighted in that week and no tick was used if the bird species was not identified. The Data was used in a project to map the distribution of birds throughout the Wheatbelt of Western Australia.

The Species Richness Index (SRI) can be calculated as a present or absent index measure. The aggregation context for this analysis is annual. As such, for the SRI, if a species appears at least once during a year, then that species gets recorded as a one (present), for that year, otherwise it is recorded as a zero. With annual aggregation, if a species is sighted every week of the year, then that species also gets recorded as a one (present), for that year, otherwise it is recorded as a zero. There are 134 possible bird species, so the maximum possible number for the SRI in any given year is 134. For each year (1987 to 2016) determine the total number of different bird species that were identified on the property, and hence determine the SRI for that year.

Conduct regression analysis for a model where the SRI is the dependent variable and Time (year) is the explanatory variable, and consider the following models: Linear-Linear, Linear-Log, Log-Log, and Log-linear, where the first term in each model expression denotes the Dependent variable and the second the Explanatory variable. For each model type, also consider the model functional form fit to the data, and whether heteroskedasticty and autocorrelation are an issue. Once you have considered each model and the relevant test select the relevant answers for Question 7.

solve using R