Data analysis of public reactions to a bear photograph on Facebook

¶ … Criminal Justice and Criminology

Interpreting simple data

The data collection exercise involved posting a picture of a bear on Facebook. The caption of the picture asked viewers to provide their thoughts on the picture. This caption was kept simple in order to prevent bias that could arise from asking viewers to like or dislike the pictures. The data that could be collected included likes, positive and negative comments. The data was collected over a 24-hour period.

The data collected can be summarized as below.

Number of respondents

Likes

Positive comments

Neutral comments

Negative comments

Gender

Total

Likes

Positive comments

Neutral comments

Negative comments

Males

Females

Level of education

Total

Likes

Positive comments

Neutral comments

Negative comments

High School

Bachelor's

From the simple data collection exercise, it is possible to determine the number or proportion of people who like the picture or post, the proportion of negative and positive reactions, the gender distribution of the responses, and the amount of time between posts. It is also possible to show the relationship between education level and the proportion of each response.

A bivariate analysis is used in the descriptive statistics. This involves using two variables to describe the data. Descriptive statistics will be able to describe the data appropriately. The data is ordinal and thus the mode is the best measure of central tendency to use.

What can be determined from the data collected?

In this post we wanted to know if the public liked bears so we chose the bivariate analysis to distinguish if two (2) variables were involved would be used in the identification of the male to female gender for the collection of the source data. We were also interesting in knowing the reactions of individuals to the picture. This is determined by collecting information on the proportion of people who like and who dislike the post.

What would be the appropriate measures of central tendency?

In this collection of data were interested in relationships among the variables. The appropriate measures of central tendency for this type of data collection are mean and mode. The mean will help to know different statistics such as the average number of males and females who posted and the average time between posts. The mode will help to know whether males or females posted more or whether the number of positive reactions exceeds negative reactions or vice versa. Additionally, we used a subgroup, to determine if a univariate analysis was appropriate so that we could find comparisons that focused on people's description and/or if there were different units for this analysis for study.

Total

In this case we also concluded that in addition to the inferential model could use statistics for computation in the suggestion of inferences a sample observation from a population that was larger. Inferential statistics in this case help us research and form our own conclusions direct from observations; and this fact finding involved conclusions about a certain age of population where the previous study drew samples from. In addition before numbers were challenged, the level of statistics used was non-toxic and safe for humans because the collection was made on social media.

Using descriptive statistics, it is possible to determine the number of each gender group who gave positive and negative responses as well as the average time between responses. Using inferential statistics, it is possible to determine whether males or females were more likely to give negative or positive responses. However when making these inferences, one would need to make assumptions that all males and all females react in the same way. It is also possible to determine whether females or males were more likely to post during a certain time of the day using descriptive statistics. The trend of postings by time can also be determined using descriptive statistics. Using this descriptive statistics, it is also possible to make inferences. One inference that can be made is on what time a specific gender is most probably to post and when more responses are expected. These inferences are also subject to the assumption that the population as a whole is likely to behave in the same way.

Conclusions that can be made

Our conclusions will include measures of central tendency. Therefore, the most frequent attribute will be either grouped or un-grouped and the arithmetic mean which will show the sum of values for all observations will be divided by the number of observations and finally the median which means the middle attribute will be ranked as the distribution of observed attributes.

The tables below show how the mode, arithmetic mean and median of grouped data are calculated. The mode is calculated by looking at the frequencies and finding the most frequent group which is this case is those aged 41 years. The arithmetic mean is calculated by multiplying the age by the frequency then adding them all up and dividing it by the total number of respondents which is 18. Since the total number of respondents considered is 18, the median or middle value is that represented by the position of the 9th value (18/2). So in this case, the median is in the 41-year-old group so this is the median value. (Maxfield, & Babbie, 2011).

Age

Number

People

Mode

Under 40

41

9

Mode = 9 (Most Frequent)

42

2

43

2

44

5

Total = 18 people

Age

Number

People

Mode

Formula

Arithmetic Mean

Under 40

41

9

Mode=9

Most Frequent

41x9=369

42

2

42X2=84

43

2

43X2=86

44

5

44X5=220

Total = 18

=369+84+86+220

Total 759 divided by 18 people

42.17

Age

Number

Median = 9 'Midpoint'

Under 40

41

9

1-9

( | V

42

2

9-11

9

43

2

11-13

44

5

13-18

9.05

Total = 18 people

The negativity score of the bear photo can be deduced by using inferential statistics. By taking a backhanded positive and negative scoring system to classify the positive and negative comments, the negativity score can be found. In this case, the type by sex or education or age mean score which is the correlation of sex to the average 'negativity score' can be calculated.

From the data collection exercise, it can be inferred or deduced that women use Facebook more than men or they are more likely to comment on posts more than men. This is attributed to the number of women who responded being more than that of men. However, this may not have reached statistical significance since the sample size was small and not powered enough. The distribution of level of education also shows that there are more people who have reached college and graduate level of education using Facebook than those in high school. This reaches statistical significance though the sample size is small. The possibility of this being a chance finding cannot be ruled out. This result is somewhat different from that of a colleague which found that teenagers who are more likely to be at the high school level of education responded in higher numbers. This difference can be attributed to the differences existing in the posts. The other post was on the use of the GPS feature on smart phones. Teenagers are more likely to engage in tech talk compared to other age groups.

Conclusion that can be drawn about interpreting data

Data interpretation is an important exercise that comes after the data is analyzed. It helps to draw conclusions from the data itself. In the data interpretation process, it is possible to tell whether the hypothesis has been accepted or not. It is also possible to know how the sampling or data collection methods affected the results that have been collected. Data interpretation also helps to provide the answer to the research question. This is because the data collected is integrated with the hypothesis in order to find an answer to the initial research question. It is also learnt that data interpretation helps to discuss the summaries and data presentation methods in order to know whether they were appropriate for the type of data collected. Predictions regarding the elements of uncertainty that result from the data collection exercise and data presented can then be discussed and analyzed to acknowledge the role that variation plays in hypothesis testing. The inferential analysis was conducted in order to help draw conclusions about the results of the small data collection exercise. The statistical tests that have been used which are the ANOVA and t-test also show the relationships that exist between the variables. Data interpretation is also an essential step towards making evidence-based predictions for future related research questions.

Reaching research conclusions

Chicago is a famously polarized city in terms of racial and class tensions, thus focusing on community needs generates an important sense of unity between members of the police force and specific community members. It has been praised for its effective implementation of POP. It recently began a much-heralded initiative to bring POP to its public school system. Killings occur with alarming regularity in the city's public schools. There were 67 deaths during the first half of 2007 -2008 academic year alone, the year that POP was instituted to combat the problem within the school system. "If this doesn't sound bad enough, the 67 doesn't include the hundreds of students who were shot or beaten but managed to survive" (Waldek 2009). To implement POP, the Chicago police first surveyed the assessed patterns of school violence and causation possibilities. Over 500 students who suffered school violence were studied. "The students at the highest risk of violence, by statistics, are most likely to be black, male, without a stable living environment, in special education, skipping an average of 42% of school days at neighborhood and alternative schools, and having a record of in-school behavioral flare-ups that is about eight times higher than the average student" (Waldek 2009). Violent outbursts also tended to occur within a relatively narrow window of opportunity, either two hours before or after school and