Skip to content

Instantly share code, notes, and snippets.

@caspercasanova
Last active October 5, 2024 02:00
Show Gist options
  • Save caspercasanova/35ceb54f5c688b7848c57cf5e933828a to your computer and use it in GitHub Desktop.
Save caspercasanova/35ceb54f5c688b7848c57cf5e933828a to your computer and use it in GitHub Desktop.

Questions:

  1. What is the general trend shown in the chart shown from the weather station at Punta Arenas? How does this support Crichton’s skepticism of global warming?

It shows that the mean tempurature has fallen since 1888 to 2004 which shows a data point validating his hypothesis.

  1. Fill in the chart and complete the blank below
Station 1951-2000 Temp (°C) 1901-1950 Temp (°C) Difference (°C) Cool Down or Heat Up
Punta Arenas 6.752 6.828 -0.076 Cool Down
  1. This data has a matched-pairs structure. Explain what that means. Then, discuss/explain why it’s useful to have paired data.

  2. Did more stations see warming during the 1900’s or cooling during the 1900’s? Discuss/explain how you can tell from this plot?

More stations saw warming and you can tell because most of the data is above zero. Zero being on the left of the graph makes it very obvious.

  1. Does Punta Arenas appear to be representative of the general trend we see in the station data? Discuss/explain how does this influence your evaluation of Michael Crichton’s evidence against global warming?

No it doesnt. If the point came within the 2.5-3 range I'd say it does since that is where the mean seems to be. It seems like Michael Crichton either cherry picked his data point or randomly guessed a point tht doesn't reflect the overall data.

Group n Mean (°C) Standard Deviation (°C)
1951-2000 (Group 1) 32 𝑥̅1 = 14.175 $𝑠_1$ = 10.985
1901-1950 (Group 2) 32 𝑥̅2 = 13.824 $𝑠_2$ = 10.910
Sample Mean Difference (Grp 1 – Grp 2) 32 𝑥̅𝐷 = 0.351 $𝑠_𝐷$ = 0.296
  • Is the sample mean difference more representative of the general trend than Punta Arenas? Discuss why or why not?

Yes because the sample size is 30 >= 0 vs a single data sampling, which makes the data more statistically significant and reflective of the overal weather stations.

  • Does the histogram and boxplot for the distribution of differences appear to be approximately normal with no outliers? Attach the images of all graphs here.

There doesn't seem to be outliers and it does seem to be approximately normal, however there if I had to be honest I'd say its almost bimodal because there are two peaks and slightly skewed left, but for all intensive purposes it is approximately normal.

image
  1. Matched-pair t-test and interval
  • Matched-pair t-test – Go back to the website and follow instructions below:
    • Use the Inference Type dropdown menu and select Significance Test.
    • Correctly select the Alternative Hypothesis direction from the drop-down menu Determine if we have statistically convincing evidence of true global warming (on average).
Null Hypothesis There was no significant change in global temperatures between the two periods.
Alternative Hypothesis There was a statistcally significant increase in global temperatures
P-value <0.0001
Decision Reject the Null Hypothesis
Conclusion statement in context Yes we have statistically convincing evidence that shows the existance of global warming in this region
  • Conclusion statement in context:

Reject the null hypothesis because the p-value is less than 0.0001

  • Matched-pair t-interval - Use the 95% Confidence Interval from the Inference Type dropdown menu. Find and interpret a 95% confidence interval for the true global average difference in temperature (Late 1900’s – Early 1900’s) in context.
image
  • Is your confidence interval consistent with your results of the hypothesis test? Discuss/Explain

Yes

  1. Should popular media portray both sides of the climate change “debate?” Discuss/explain your reasoning. Not sure what is meant by "Both" sides of a debate. Show us the data, show us the process of getting that data, show us your math and if you make a conclussion, show your evidence, otherwise let the people refute your claims. If data is intentionally midhandled or mis represented, we should drag those people through the mud. Harvard Data Scandal

Media doesn't have a scientific approach to accepting critisms.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment