In the video, Professor Hans Rosling demystifies statistics turning the stream of data into useful information. His interpretation of data, graphs, and descriptive statistics makes it easy to understand statistics as a practical subject and not theory. . For example, he takes a real life situation, assigns statistics values to the real life information, and conducts a descriptive statistical analysis to provide the audience with relevant inflation. Then it become very clear that data can be used to analyze all kinds of social problems and find a solution to these social problems from hunger, crime, social economic, problem, traffic ad all these information aid in decision making which is further enabled by data analysis tools. These analyses can be done in real time therefore helping solve social economic problems in time.
I also realized the important part played by the community without their knowing. The citizens stream data in real time though the social media and the internet without knowing. It is interesting realize that instead of the government watching the citizens, the citizens can also watch the government and make them accountable. The power of statistics can improve government and citizen austerity and it provide useful insight based on statistical interferential. The speaker has clearly made business case, and the realization that statistical is not only useful in specific sectors is a new awakening as I have realized that there is a lot more to practical statistics as opposed to theory.
2. Short Videos:
In this video, Ben Goldacre unpicks dodgy scientific claims made by scaremongering journalists, dubious government reports, pharmaceutical corporations, PR companies and quacks. He takes bad science head on and debunks how the people in the know manipulate data to produce biased reports. For example, he argues that people use their authority backed by lots of certificates to make believe their statistical information. Bias by industry. The presenter makes it clear that most scientific data have been manipulated to meet the interest of the special interest groups and may not be for the benefit of the Society.
David McCandless refutes the argument about data glut because to him, information overload is beautiful if the data can be visualized. I was amazed with his version of data visualization when he introduced the world debt and realized that all disciplines can be visualized and this is the same idea that the media use to create fear in man when they visualize data incorrectly. Data visualization make data come to life and this can be done to data of all types to make understanding data easy and faster.
Joel_Selanikio halps in understanding the main problem of big data because to him big Data is compel and may have allot of errors and this affects data quality because little data, old data and missing data is analyzed and the decision are based on these low quality data. However, data can also be of low quality if it is received later and the results are of no use, but technology as enabled real time data capture analysis and reporting thereby improving the usefulness of data.
Anne milgram argues that the old school of fighting crime is not realistic, is not effective and additionally time consuming. The criminal justice system, police departments, prosecutors office, the court ads well as the jails failed to be effected because their decision were not data driven. To anne_milgram, smart data, rigorous statistics and analysts can be unused to, solve criminal justice problems such as violence, prosecuting string gland, gun and drugs trafficking.
3. Chapter 17::
1. Define what non-parametric statistics are. (5 points)
Non-parametric statistics refers to the statistics involving the analysis of raked data. For example, this statistics is used when some assumptions about a specific dataset is not clear and so questionable. One of the key test statistics is the Wilcoxon Mann-Whitney test that assumes that the difference between two samples is normally distributed. nonparametric statistics is referred to as distribution free test because it does not depend on any precise assumption about the data’s distribution. There are no clear parameter estimates for the nonparametric statistics
What does a chi-square test do? When would you use it? Provide an example not mentioned in the text where a chi-square test would be appropriate. (10 points)
This is the test used to compare the observed categorical data and the expected categorical data. This trust statistics is particularly useful when analyzing unpaired data and the sample is very large. The chi square test is specifically used to compare data test by carrying out the tests of goodness of fit and tests of independence. An example of chi square in practice is when determining if the race of a person determine his response to economic development. In this case, the question is interested in positive responses about economic development.
Frequency of positive response
American Indian and Alaska Native
Native Hawaiian and Other Pacific Islander
The null hypothesis may be Ho: there is as significant relationship between the race and frequency of positive response
3. Is determining a critical value for a chi-square similar or different from determining a critical value for a parametric statistics? How so? (5 points)
4. In a chi-square test, what is an expected value? (5 points)
The expected values in any chi square test include all the values that would be expected under the null hypothesis of no association. Expected values can be calculated from the data by determining the expected values for each cell. For example, to determine the expectative value for each of rah cell in the two way table, it is advisable to multiply row total Colum total the divide by the total number of observations (RXC)/n)
5. Report the results of the tests in videos 17.4, 17.5, 17.6, and explain in your own words what they mean. What is the error message that appears on screen in video 17.5? (5 points)
4. Reflection: Reflect upon the learning process via video in this chapter to that of the online modules and that of traditional lecture format
Throughout this course, I realized that learning via video is relatively interesting and convenient as it is easy to watch the video anywhere and on any gadget. For example, I loved the fact that I could watch the video on my tablet and smart phone even when I am on the move making learning virtually portable. On the other hand, online learning modules requires one to have internet access despite being convenient especially for those who work and live far from the university like me. However, it may be difficult for me especially when I am in a remote area for function where internet access is not available. I would prefer using the online learning method because it is easy to conduct research to backup the online course tutorials. Finally, the traditional lectures are much more preferable than the other learning methods because it can be engaging especially considering the fact that other students and the lecturer are around to ask questions and get answers in real time. It is also easy to get faster responses and insight from other students that are not the case when learning via video and online method. However, traditional learning method is time consuming, very expensive considering the traveling expenses and inconveniencing. Given an opportunity to chose, I would chose convince, and quality learning making online learning more effective as compared to the traditional learning methods or learning via video.
Ben Goldacre. (2011, November 21). Battling bad science. Retrieved April 13, 2014, from https://www.ted.com/talks/ben_goldacre_battling_bad_science
Selanikio, J. (2013, February 20). The surprising seeds of a big-data revolution in healthcare. Retrieved April 13, 2014, from https://www.ted.com/talks/joel_selanikio_the_surprising_seeds_of_a_big_data_revolution_in_healthcare
Anne, M. (2013, October 2). Why smart statistics are the key to fighting crime. Retrieved April 13, 2014, from https://www.ted.com/talks/anne_milgram_why_smart_statistics_are_the_key_to_fighting_crime
McCandless, D. (2010, July 21). The beauty of data visualization. Retrieved April 13, 2014, from https://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization