RSDCS assignment: Data analysis and science communication

RSDCS assignment: Data analysis and science communication

Alice Risely, a.risely@salford.ac.ukNovember 2023

Part 1: Data analysis (60%)

For this section you will need conduct five data analysis exercises that aim to assess your ability to handle, plot, and statistically analyse a biological dataset. Make sure all your answers are written out using complete sentences, do not simply copy and paste outputs of statistical models or add screenshots of your outputs. Where specified, copy and paste your plots under the relevant questions. All plots need to have neat axis labels and, where specified, a title. Plots can be coloured and themed however you wish, you will not lose or gain marks for how the plot looks.

For exercises where you are required to identify the correct statistical model, you will need to present 1) the model name, 2) the effect size, and 3) the p value. For p values under 0.001, you can simply put p < 0.001. For all other p values, put the exact figure to three decimal points (e.g., p = 0.025). Note: You will get outputs in scientific notation (e.g., 1.23e-5). 1.23e-5 is equal to 0.0000123 (move the decimal point five places to the left).

The dataset: A study on the breeding ecology of the blackbird (Turdus merula)

You are a wildlife biologist interested in identifying how urbanization might impact bird breeding biology. To investigate this question, you conducted a long-term field study on the European blackbird, Turdus merula. Between 2015 and 2018 your research team captured male blackbirds in early spring across urban and forested sites and marked them with plastic colour rings on their legs and a small radio transmitter on their tail so you could track them across the rest of the breeding season. When you captured them you collected data on their body condition (a proxy for how much food they have access to) and their age category (adult or sub-adult). After you released them, you collected data on how long they spent singing in the morning, and how many chicks they successfully produced over the entire breeding season.

After a lot of hard work, you had full data for 200 male blackbirds! This dataset in available to download alongside this assignment on Blackboard. Your dataset looked like this:

Below is an explanation for each column and whether it should be treated as catagorical, continuous, or count data.

Column name Explanation Data type

blackbird_id Unique identifier for each tagged blackbird Categorical

year The year the bird was captured and tagged Categorical

habitat Whether the habitat was classified as urban and forest Categorical

site The location the bird was caught Categorical

body_condition A measure of body mass index, accounting for wing length. Body condition represents access to food resources Continuous

age_category Whether the bird was adult or sub-adult (based on plumage) Continuous

song_length_mins Number of minutes per hour spent singing after dawn, measured as the average from four dawn surveys. Continuous

number_offspring The number of chicks successfully fledged over the breeding season Count

Exercise 1 (12 marks)

What was the mean and minimum number of minutes birds spent singing in the morning? (2 marks)

What was the mean and maximum number of offspring that blackbirds produced? (2 marks)

How many blackbirds were captured in forest habitats, and how many in urban habitats? Write your answer and in addition present a barplot. Make sure you add a title and your x axis label is capitalised (5 marks)

You want to check if there is any difference in the number of birds caught between forest and urban habitats. Conduct a chi-square test to test this, and summarise your conclusions in one sentence, providing statistical support from your model. Do not just copy and paste your statistical output, but embed the results within your written conclusions (3 marks).

Exercise 2 (16 marks)

When food resources are low, birds spend more time trying to find food and less time singing to attract a mate. However, when birds have easy access to food, they spend more time singing to attract a mate. Bird feeders in urban gardens provide a reliable food source for birds, and you suspect that urban blackbirds will therefore be able to spend more time singing to attract a mate. You want to test this theory statistically.

Write out your null and alternative hypotheses. Remember, hypotheses are explanations for observed phenomena, not just predictions (4 marks).

Examine whether time spent singing has a normal (bell-shaped) distribution. Write 1-2 sentences justifying your conclusions. In your answer, refer to the results of a Shapiro-Wilk test of normality, and include a histogram. Make sure that your x-axis label is informative (4 marks).

Present a plot of the relationship between how long blackbirds spend singing and habitat. Make sure your axis labels are informative and include a title. Colour the plot how you like (4 marks).

Conduct the correct statistical test for you data, based on your answers in b and c.Write 2-3 sentences on your conclusions, providing statistical support from your model. Do not just copy and paste your statistical output, but embed the results within your written conclusions (4 marks).

Exercise 3 (18 marks)

If urban blackbirds have better access to food than forest blackbirds, then you also reason that perhaps they might successfully produce more offspring each breeding season. You want to test this statistically.

Write out your null and alternative hypotheses. Remember, hypotheses are explanations for observed phenomena, not just predictions (4 marks).

Is number of offspring normally distributed? Write 1-2 sentences justifying your answer. In your answer, refer to the results of a Shapiro-Wilk test of normality, and include a histogram. Make sure that your x-axis label is informative (4 marks).

Present a plot of the relationship between the number of offspring blackbirds produce and habitat type. Make sure your axis labels are informative and include a title. Colour the plot how you like (4 marks).

Conduct the correct statistical test for you data, based on your answers and plot in b and c.Write 2-3 sentences on your conclusions, providing statistical support from your model. Do not just copy and paste your statistical output, but embed the results within your written conclusions (4 marks).

What does this p-value represent? Explain in one sentence (2 marks).

Exercise 4 (14 marks)

Birds that spend more time singing are more likely to attract a mate, and therefore may be more likely to successfully raise more offspring. You want to test this theory statistically with your blackbird dataset.

What are your null and alternative hypotheses? (4 marks)

Present a plot of the relationship between the time blackbirds spend singing and the number of offspring they successful fledge. Include a linear trend line. Make sure the axis labels are neat and informative, and include a title (5 marks).

Conduct the correct statistical test for you data. Note that you have already tested both variables to see if they are normally distributed so you do not need to repeat this. Write 2-3 sentences on your conclusions, providing statistical support from your model. Make sure you include in your answer the direction of the relationship (positive/negative). Do not just copy and paste your statistical output, but embed the results within your written conclusions (5 marks).

Exercise 5 (16 marks)

You want to understand which environmental and biological variables predict time spent singing. Fit a general linear model (GLM) with the time spent singing as your response variable and habitat, body condition, and age category as your three explanatory variables.

Present a table of your model results, including predictor estimates, standard errors, effect size, and p values (5 points)

What is the R2 of your model? What does this tell you about the explanatory power of the model? (2 points)

Write 3-4 sentences on your conclusions, providing statistical support from your model. Your conclusions should include information on the strength and direction of the effects on time spent singing. Even though you have provided a table with the statistical output above, you still need to include the relevant effect sizes and p values in your conclusions (9 marks).

Part 2: Science communication (40%)

Write a 300 to 350 word summary on your findings regarding the relationship between habitat and blackbird behaviour and breeding success. Based on your results, how might urbanization impact blackbird populations? What are the potential mechanisms? Are there some additional mechanisms that could account for your results in exercise 5 that the study didnt account for?

Your summary should include 2-3 sentences each on the following topics: 1) the background knowledge on this subject, including 3-4 scientific references; 2) your aims and/or predictions; 3) Your methods and results; and 4) your conclusions. At the end of the summary, you should include a references section with the correctly formatted references (in APA 7th style). It should be no more than 350 words. The summary will be marked according to the standard Level 7 grading descriptors that can be found in the Assignment folder.

Note, there is no right or wrong answer for how you embed the results of your study in the context of wider research, although your conclusions must be based on the evidence generated during the analysis. You also should not include statistics in your summary.

Youre references can (but dont have to) include:

Ruffino et al.2014. Reproductive responses of birds to experimental food supplementation: a meta-analysis. Frontiers in zoology. https://doi.org/10.1186/s12983-014-0080-yKempenaers et al.2010. Artificial night lighting affects dawn song, extra-pair siring success, and lay date in songbirds. Current biology. HYPERLINK "https://doi.org/10.1016/j.cub.2010.08.028" o "Persistent link using digital object identifier" t "_blank"https://doi.org/10.1016/j.cub.2010.08.028