Survey Project

Hypothesis: Rstudio and Google will yield more “random” (have a greater variation) coin flips than humans.

Importance: This research is important because when conducting studies subjects are often randomly chosen, and if for some reason these individuals are not randomly chosen it can lead to skews forming in the data. This research suggests that using a computer program to select individuals may lead to a more “random” pool of subjects than human selected pools of individuals and would therefore infer less skewed data.

Methods: First, 10 samples of 10 coin flips from Rstudio, Google, and (10 random) human individual subjects were collected. Than the average % of heads and tails for each group were found. Than Excel was used to find the variations for each group to see which group had the most variation from the mean and was therefore the most “random”.

R-studio coin flips:

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "heads" "tails" "tails" "heads" "heads" "tails" "heads" "tails" "heads"

This sample is 60% heads, 40% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "heads" "heads" "heads" "tails" "heads" "heads" "heads" "heads" "heads"

This sample is 90% heads, 10% tails

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "tails" "heads" "tails" "tails" "tails" "tails" "tails" "heads" "heads" "heads"

This sample is 40% heads, 60% tails

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "heads" "tails" "tails" "tails" "heads" "heads" "tails" "tails" "tails"

This sample is 50% heads, 50% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "tails" "tails" "heads" "tails" "tails" "tails" "tails" "tails" "heads" "tails"

This sample is 20% heads, 80% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "heads" "tails" "heads" "tails" "tails" "heads" "heads" "heads" "tails"

This sample is 60% heads, 40% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "tails" "heads" "tails" "tails" "heads" "heads" "heads" "tails" "heads"

This sample is 60% heads, 40% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "tails" "heads" "heads" "heads" "heads" "heads" "heads" "heads" "tails"

This sample is 70% heads, 30% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "tails" "heads" "heads" "heads" "tails" "tails" "heads" "tails" "heads"

This sample is 60% heads, 40% tails.

coin <- c('heads', 'tails')

sample(coin, size = 10, replace = TRUE)

 [1] "heads" "tails" "heads" "tails" "heads" "tails" "heads" "heads" "tails" "tails"

This sample is 50% heads, 50% tails.

Google samples: 1. ttththhtth 40% heads, 60% tails. 2. thhhthhtht 60% heads, 40% tails. 3. ttthhhttth 40% heads, 60% tails. 4. hhttthhtht 50% heads, 50% tails. 5. tthhtttthh 40% heads, 60% tails. 6. thtththtth 40% heads, 60% tails. 7. hhhthtthht 60% heads, 40% tails. 8. hhtthhhttt 50% heads, 50% tails. 9. hthththhhh 70% heads, 30% tails. 10. thhhthhthtt 50% heads, 50% tails.

Human samples: 1. hhthtththh 60% heads, 40% tails. 2. thhhtthhht 60% heads, 40% tails. 3. htthhhttht 50% heads, 50% tails. 4. thhhttthtt 40% heads, 60% tails. 5. hthhtthhht 60% heads, 40% tails. 6. tthtthhthh 50% heads, 50% tails. 7. hhhththtth 50% heads, 50% tails. 8. hhththttht 50% heads, 50% tails. 9. hthhtthtth 50% heads, 50% tails. 10. ttthhthhht 50% heads, 50% tails.

Results: R studio produced 56% heads and 44% tails on average with a variation of 17.436. Google on average produced 50% heads and 50% tails with a variation of 10. Humans produced 52% heads and 48% tails with a variation of 6. These results support the hypothesis that humans are the least “random” and that Rstudio and Google are more “random” in terms of coin flipping. Rstudio variance>Google variance>human variance

Next Steps: Further research should be conducted to see if this pattern continues. Future studies should look into if other programs produce similar or greater variance, if the variance increases or decreases with the amount of coin flips that are done, and if this could be advanced to simulate the selection of individuals for studies.