这个作业是用R语言完成LOGIT模型和计数数据模型等
SCIE4402
Assessment Guide for Day 4 Stream B:
surveys, binomial and Count data
General Marking matters
The questions test general competence in estimating models and explaining results. The level of difficulty and guidance provided varies with each question. Presentation quality, and the quality of the writing will also be assessed. If the question asks for a visual plot make an effort to get the plot looking nice.
Assignment Due Date
As per the unit outline.
Assignment Question 1: logit model (zero, one data)
The relationship under investigation is the choice of possums to eat feed type 1 or feed type 2. The variables record the distance each food source was placed from the possum nest (in metres), and records the choice of food type (0 = food type 1 and 1 = food type 2). The data is in the Feed1 data file. The research question you are trying to answer is, if a possum is located equal distance from each feed type, will they choose feed type 1 or feed type 2?
Your answer must fit on a single page and must include: (i) an appropriate plot; (ii) a regression results summary table; and (iii) a statement about which feed we expect a possum to choose, along with sufficient evidence to support this conclusion. (Assignment question 1 is worth 5 marks.) Note: The text required to answer this question is not substantial, just a sentence or two. The main assessment element for this question is the quality of the plot and the quality of the summary table.
Assignment Question 2: RECREATION demand (count data model)
The RecreactionDemand3 data file contains information from a survey of people about how frequently they have visited Perth Park in the past 12 month. The costpark variable is the estimated travel cost for people to get to the park. The costsub variable is the cost of getting to an alternative park that has the same features as Perth Park. The income variable is a measure of the respondent’s income.
Background
From independent records it is known that 80,000 people visit Perth Park each year. The government is considering increasing the cost of visiting the park by $5. Before the government makes a decision the minister wants to know certain information. Specifically, the minister wants to know the following things:
- for a person faced with average income and costs, how often are they likely to visit Perth Park at the moment, ie what is the predicted number of visits at the moment for such a person, and following the increase in price, what is the predicted number of visits for such a person.
- Given 80,000 people visit the Park each year, using a discount rate of 3 percent, 5 percent, and 9 percent, and the current information on the cost of visiting the park, what is the Net Present Value of the recreational value of the Perth Park to the community?
- The shape of the distribution of Park users described via a histogram.
- Are there any important qualifications to the results.
Your answer must fit on a single page. The text required to answer the questions is not substantial. When presenting answers to ministerial questions you do not need to provide any technical detail, but what you have done needs to be clear. This means you need an introduction sentence that provides details on the issue your results address. To answer question part (i) you can use either one or two sentences to state the outcomes. Similarly, for question part (ii) you can write a single sentence or use a table. Note if you use a table it must be formatted to journal publication standard. For part (iii) you may need to go back to look at material from earlier in the course to refresh your memory on the creation of a histogram. Again, for full marks the quality of the plot must be publication standard. For part (iv) you just need to think about what you have done. (Assignment question 2 is worth 5 marks.)
Assignment Question 3: Hedonic pricing (continuous datA)
The hedonic data file contains information about house sale prices and the characteristics of these houses. The data columns are as follows:
House price = is the house sale price ($’000)
bedrooms = the number of bedrooms for the sold house
bathrooms = the number of bathrooms for the sold house
land area = the land area of the sold house (m2)
house age = the age of the sold house
green space = whether the sold house located within walking distance to public open space
time = year of the house sold
location = the suburb the sold house located
Prepare a short written report (no more than two pages including tables) that investigates the implicit price of living within walking distance to public open space and also explains your approach to modelling. Hint: you will want to consider as part of your testing approach: (a) a spatial fixed effects model; (b) a time fixed effects only model; and (c) a time and spatial fixed effects model, and use formal tests to establish the model you think is most appropriate. You will also want to include a summary table for the final model you estimate, and reference a method, such as the box-cox transformation to help guide your selection of model functional form (transformation of the Y variable) (Assignment question 3 is worth 5 marks). Relative to Question 1 and Question 2, which are quite straightforward, this question involves more work. The data is not ‘well behaved’ so you need to make judgment calls about what to do. A wide variety of approaches are acceptable.
Assignment Question 4: Willingness to pay
Context:
The Ministry of the Environment of Japan is considering creating a new Marine Protected Area (MPA) in Hokkaido. MPAs are to ensure “the protection of the places of national scenic beauty and also, through the promoted utilization thereof, at the contribution to the health, recreation and culture of the people and ensuring the conservation of biodiversity”.
It will however be costly to maintain the new MPA and so the Ministry has conducted a survey to gain information on community Willingness to Pay for this protected area. Specifically, the Ministry has conducted a survey and asked a representative sample of Hokkaido residents if they were willing to pay different values to ensure the creation of a new MPA. Depending on the answer to the first WTP question, respondents were then asked a follow-up question. The survey data is attached, and in addition to whether or not a respondent said yes or no, information was also collected on respondent Age, Gender, and Income. In the data set respondents were asked only to indicted age and income only in broad groupings. In the data set higher numbers for income indicate higher household income, and higher numbers for age indicate older respondents. The data file is the HU_out file and the specific variable descriptions are below:
Bid_1 – first WTP value (values are ‘00s of yen)
Bid_2 – Second WTP value (values are ‘00s of yen)
Ans_1 – response to the first question 0= no, and 1 = yes
Ans_2 – response to the second question 0= no, and 1 = yes
Gender – M is for Male and F is for Female
Age – higher numbers = older respondent
Income – higher numbers = higher household income