Notes

if a type I error occurs, we reject the null hypothesis, but the null hypothesis is true. This means that we will find our null hypothesis which states that p1-p5 = the true proportions of people who range from strongly disagreeing to strongly agreeing to be false, and instead prove our alternative hypothesis that p1-p5 do not equal these true proportions. this is false, and this type i error then means that our data is less significant than we would assume

if a type II error occurs, we fail to reject the null hypothesis, but the null hypothesis is not true. This means that we will find our null hypothesis which states that p1-p5 equal the true proportions of people who range from strongly disagreeing to strongly agreeing to be true, thus disproving our alternative hypothesis. this is false, and a type ii error in this situation could result in us assuming our data is less significant, when it actually is more significant than the level of significance

in either situation, a type i or type ii error could be detrimental to the analysis of our data, and potentially misconstruing the true implications of word choice bias in influencing people's responses as well as the effects on morality

we have concluded that there is a significant difference between our observed and expected values, aka our values from survey 1 and survey 2, respectively, for each of our six questions. this means that our survey results are significantly difference, and we can say that word bias likely plays a role in influencing responses since the responses were so clearly significantly different. additionally, on a less-statistical note, human morality can clearly be shaped through the influence of bias when provided with different information or using different words`, indicating that it is not a concrete construct but rather an everchanging perspective on issues.






once we collected the data, we constructed bar graphs and determined relevant descriptive statistic values for each situation.
here we can see that this data is skewed to the left, with both the greatest number of people and the median number of people who selected option 4, which means they slightly agreed with the statement that the natives should remain isolated.
for the second situation, the data is also negatively skewed, and the median of the data is 4 whereas the mode is found to be 5, so most people either slightly agree or strongly agree that the heart should go to the child
in this third situation, the data is left-skewed, and the mode and median are both again 4, so people largely selected option 4, indicating that they slightly agreed that the treatment using CRISPR gene editing is ethical
the fourth situation discussing population control policies has a negatively skewed bar graph, and the median and mode are both 4, so most people slightly agree that countries should enforce population control policies
the fifth situation depicts a robin-hood like scenario, with, again , left skewed data and both a median and mode of 4, which shows how many people slightly agree that you should not inform the police
in the sixth and final situation entailing the trolley problem, the data is skewed to the left, and the median and mode are both 4, so most people slightly agree you should not pull the level

In the second survey, the same situations have been used, but they have been described using very contrasting word choices and statements.
question 1 has a negatively skewed bar graph with a median and mode of 4, so most people slightly agree that the island should be explored
in the second situation, the data is again skewed to the left, but the median is 3 and the mode is 4, so people are generally either neutral or they slightly agree that the heart should go to the elderly citizen
the third situation again discusses CRISPR technology, and the bar graph shows the negatively skewed chart with a median and mode of 4, so most people slightly agree that CRISPR treatment is unethical
in the fourth question, the again left skewed data and median and mode of 4 shows how most people slightly agree that countries should not enforce population control policies
the fifth situation shows a graph that is roughly skewed to the left. the median is instead 3, and the mode is 4, so most people either have no opinion or slightly agree that you should inform the police of the theft
in the sixth and final question, the data is again slightly skewed to the left, the median is 3, and the mode is 4, demonstrating how people are either neutral or slightly agree that you should not pull the lever in response to the trolley problem

when performing our hypothesis tests, the P1 through P5 values shown at the top are supposed to be written down as a decimal by dividing each of the numbers from 44, so the true values for the null and alternative hypotheses of p1-p5 have been included on the side. for each of the six tests we must conduct, p1 will be the proportion of people who strongly disagree with a given statement, p2 will be the true proportion of people who slightly disagree, p3 is the true proportion of people who are neutral, p4 is the true proportion of people who slightly agree, and p5 is the true proportion of people who strongly agree. the conditions that the sample was taken randomly and all expected counts are greater than or equal to 5 remain constant. lastly, the observed counts are the values of the number of people who selected from strongly disagree to strongly agree from survey 1, and the expected counts are the values of the number of people who selected from strongly agree to strongly disagree from survey 2. survey 2 proposed an opposite statement to that of survey 1 (such as this treatment is ethical vs this treatment is unethical), so the values need to be switched. strongly disagree will serve as strongly agree and vice versa, so the values can be compared accurately since strongly agreeing with one statement is essentially the same as strongly disagreeing to its opposite.