Friday, March 21, 2014

Misleading Statistics (Term 1, 2014 Monday Wednesday Friday 6:00 PM Class)


            Misleading statistics, which are false data based on studies, are used to confuse people (Cambridge advanced learner’s dictionary). The first statistic was published in 1663 (Wikipedia). Since then, statistics have been widely used by millions of people all around the world in many cases including business, education, advertisements et cetera. However, statistics can be misleading, so they should be interpreted with caution. This essay aims to demonstrate various strategies that can make statistics unreliable such as sampling bias, misleading graphs, misleading advertisements and unfair comparisons.



Australian Centre for Education
Writing Assignment
           
Level: EAP5                                                                   
Day: Mon/Wed/Fri
Time: 18:00-19:30

Sampling Bias
Response bias usually creates a misleading statistics. Due to its procedure, which is applied by interviewers to collect data from participants in the survey. The information is usually not trustworthy because of two main factors. Firstly, interviewers are questioning the people face to face. So the intention of participants to avoid humiliation, the participants are likely to provide false information. Secondly, completing a questionnaire publicly causes the information to have no confidentiality. Therefore, it is unreliable data as well (Sternstein, 2010). We cannot obtain the true information from the following survey such as the level of girls personal hygiene, a doctor are not revealing how confident he is to cure a disease, and a couple seemingly lie about the problems that they have in their married life.

Healthy user bias is also one among the other factors that could make a misleading statistic. It means that they randomly pick only healthy people to do the test and the result will be the data, which will be used to represent the whole thing. Normally, the first reason is when researchers want to do research on how healthy people are, the researchers aim to go somewhere crowded such as a company, an office, or a supermarket, and so forth. They will randomly select a number of people.   Normally, people, who are picked by the researchers there, are healthier than people who stay at home (William, H. S., Amanda, R. P & M, A. B, 2011),because sick people may stay at home to get a full rest, and take medication. Secondly, if they want to lie about the information on how easily people would be at risk of diseases, researchers would select only old people to the do survey because they are lack of vaccination (Sturmur, n.d). For example, if they want to exaggerate about the rate of cervical cancer, they will select only old women to do the survey. Third, in case of doctors aim to study about the disease’s symptom, while patients have been using many different drugs, so the result might be inaccurate because that the symptom might not appear in the period of medication. Furthermore, the result might be shown as a false negative, which means those that do have diseases and tests are shown as negative (Jo, Raquel et al., 2009) For instance, infectious diseases might not have a fever as they might have used the anti-fever drugs.
            Self-selection is another method that can lead to an invalid sample. It can be defined as a kind of sample that is uncertain and biased, and which is a serious problem leads to a misleading statistics due to the small sample size, voluntary sample, the accessibility of the participants and the area where the sample is obtained. Respondents of a survey have an option to choose whether they would like to take part in the survey or not. This can make an unequal and unacceptable result. In some cases, people with a strong opinion regarding the topics will be more likely to participate. The sample can be biased due to geographical and technological development differences. For example, a telephone sampling includes only people with cell phone, so the other people cannot participate in this kind of surveys, especially in remote areas where the majority of the population over there cannot afford to own a cell phone (Hudson, Seah, Hite &Tim, 2004). Another example is the Internet polls, where only the Internet users or people from the areas where Internet access is available can take part. As a result, theo btained sample can be inaccurate, and cannot represent the whole population needed in the study. Moreover, another problem within this sample is the attempt to manipulate the result of the survey by a particular group of people. An example can be seen in the voting of Book of the Year Award in the Netherlands in 2005, which the winner was the Bible translation due to a movement by the Christian society. Therefore, this group of people cannot represent the whole population of the Netherlands. (Bethlehem, 2008).

Exclusion bias is one of the problems that occur in statistics in which someone or something does not get included in a study or data. In this case, let’s take a look at some imaginary examples-a group of media students plan to do a survey in X area base on the amount of people in the area that enjoy the media such as watching films, listening to broadcast news or reading newspapers and magazines. However, when the research is done, they find out that five families are not included in the data because they just moved out of the X area, although their names remain listed in the registration. Therefore, the five families are stated as the subject of exclusion bias. Moreover, another error might occur when individual Z happens to be in the X area for several days, but haven’t gotten their names listed in the registration. In both moving-in and moving-out cases, they can be listed as exclusion bias.














Bibliography:
1.      Sternstein, M. (2010). Barron's AP Statistics. (p. 169). New York: Barron's Educational Series, Inc.

2.      William, H. S., Amanda, R. P & M, A. B. (2011). Healthy User and Related Biases in observational Studies of Preventive Interventions: A Primer for physicians. J Gen Intern Med, 26 (5), 546-550.
3.      Sturmer, Til (Unknown). Lesson about Confounding and Selection Bias Taught by the New User Design [PDF file]. Available from: http://www.pharmacoepi.org/pub/1c22d3d0-2354-d714-5186-0d8a16a5ffa6[Accessed 11February 2014].
4.      Jo, Raquel, et al., (2009). Clinical tests. A Systematic Reviews. Available from:https://www.york.ac.uk/inst/crd/SysRev/!SSL!/WebHelp/2_2_DIAGNOSTIC_TESTS.htm[Accessed 9February 2014].
5.      Hudson, D., Seah, L., Hite, D., &Tim, H. (2004). Applied economics letters. (4 ed., Vol. 11, pp. 237-240). Retrieved from http://www.tandfonline.com/doi/abs/10.1080/13504850410001674876[Accessed 9February 2014].
6.      Bethlehem, J. (2008). How accurate are self-selection web surveys? Retrieved from http://www.cbs.nl/NR/rdonlyres/EEC0E15B-76B0-4698-9B26-8FA04D2B3270/0/200814x10pub.pdf[Accessed 9February 2014].




Misleading Graph

            Statistics are used in almost every aspect of people’s everyday lives and play an important role of illustrating data. People tend to explain most complicated figures into graphs so that it is much easier to understand. However, the information can be exaggerated, and it misleads intentionally or accidently in many ways. Misleading graphs can deceive people and lead them into a misunderstanding perspective of the charts. Frequently, scientists, experts, or researchers, seem to use misleading graphs to prove their work in a significant way. According to a statistic expert, the writer of the book which relates to the introduction of statistics said that the disingenuous graph is constructed in an inaccurate portrayal or any confusing ways just in order to covey the wrong ideas to the users (Kirk, 2008, Internet).The aim of this essay is to demonstrate the misleading graphs in various ways such as scaling and axis manipulation, biased graph and 3-dimensional graph, by supporting with the example in order to provide more accurate ideas about deceptive displays. 
            First of all, the most common use of misleading graphs is scaling and axis manipulation. This sort of graph works along with vertical and horizontal axis; it usually has the same scale in each unit so that the graph looks unique and reliable (Utts, 2005, Internet). Nevertheless, the way of using this type of display may result in deceptive information and failure. For instance, if we talk about the stock market of a company that has a slight decrease or increase, the graph will show almost a steady line with no fluctuation. Thus, they may increase the proportion either on X or Y-axis intentionally in order to make it look like it has decreased or increased more in a significant way and attracts the viewers. This kind of graph is also called the gee-wiz graph (Haney, 2011, Internet).

Figure 1: Stock Price
Moreover, truncated graphs can also result in misleading information. Regarding the following graph, it can be noticed that the average house price does not start form “0” on Y-axis and the bar chart demonstrates the average house price is £80,000 and £82,000 in 1998 and 1999.Without paying much attention on it, we may assume that the price trebled within only one year, but in reality, it does not. It is just a slight change in the way of displaying the graph, and it can absolutely make people have a biased perspective about the information .Furthermore, if the truncated bar graph is similar to the old one, and a part of graph is cut so that the short bar graph will probably appear a little from x-axis (Rensbergers, 2009, Internet). That kind of graph can give information in wrong ways and misleading data, since it may show an extreme difference of the two graphs. In fact, it is the only way they were trying to show the same figure, but in the artificial way in order to make people have a new confusing idea about it.

Figure 2: House Price

            The second most used misleading graph is the biased graph. It includes the use of two or more different sizes pictures. Although there are two different sizes pictures with the same amount of data, people tend to put their eyes on the bigger size onerather than the ordinary ones. Some graphs are made in this technique to deceive viewers to believe that the result of sales, works, annual incomes, stock markets or even the recycle products increase dramatically over some periods of time.For example,the following picture shows that the trash is enlarged significantly over a short period. It means that the same pictures are magnified to make a higher height and a wider width. If we look at the area of the picture and compare the trash in 1960 and 1980, we may think that the trash increased 3 times while the real one was about two times only (Haney, 2011, Internet).This type of chart is one of the most disingenuous displays that people usually do, since it will probably enable the maker to convince the viewers to believe in the inaccurate information. 
           
Figure 3: Trash Amount

            The third of misleading graph is 3-dimensional graph, which is the most attractive and beautiful designed graph, as it gives people the ability to see the picture in 360 degrees (Rumsey, 2010, Internet). In contrast, advantages and disadvantages always come together.The more authentic the graph is, the more misleading it becomes. From the following picture, it can be considered that there is no scale on vertical axis and the sales in 1995 are much higher than the sales in any other years. In fact, it is identical to 1997 if we see it in 2-D graph.


Figure 4: Number of sales from 1995-1998


Last but not least, in 3-D pie chart below shows that the most numerous pie is D followed by item B, then item C and the least is item A. In contrast, this information is totally wrong from what we can see in the real regular pie chart. Item D is 42% which is the same as item B. However, the 3-D pie expands the front picture so that it looks bigger and the back one becomes smaller.

Figure 5: 3D and 2D pie charts
To sum up, all of these three types of misleading graphs are what people should be careful of, as we may be tricked bythem every time in everything that we are living with.  Misleading graphsdo not only trick its viewers, but also spins their heads around to misunderstand the real things.
























Biography

1.      Haney, B.R., (2011). ‘How to Lie With Statistics.’Math 143 Project[Online]. (Springed),Calvin College.Chapter 5, 6. Available from:http://www.calvin.edu/academic/economics/faculty/bios/HaneyDocs/ [Accessed 1 February 2014].
2.      Kirk, R.E., (2008), ‘Frequency Distributions andGraphs’ [online]. Statistics: An Introduction. (5thed). Belmont, USA, Thomson Higher Education. Pages 52-54. Available from: http://books.google.com.kh/books?id=W4t9Nfgk03AC&printsec=copyright&source=gbs_pub_info_r#v=onepage&q&f=false [Accessed 4 February 2014].
3.      Rensberger. B., (2009).  ‘Slanting the Slope of the Graph.’ The Washington Post [Online] 10 May. Available from: http://www.highbeam.com/doc/1P2-831228.html [Accessed 3 February 2014].
4.      Rumsey, D.J., (2010). ‘Ten Common Statistical Mistakes.’ Statistics Essentials for Dummies [Online]. Indiana, USA, Wiley Publishing Inc. Pages 155-162. Available from: http://books.google.com.kh/books?id=QBmsVY0p7YkC&pg=PA155&redir_esc=y#v=onepage&q&f=false [Accessed 2 February 2014].  .
5.      Utts, J.M., (2005). Seeing through statistics [Online]. (3rded). Belmont, USA, Thomson Brooks/Cole. Pages 146-147. Available from: http://books.google.com.kh/books?id=j5xWsf4DD58C&printsec=frontcover&source=gbs_vpt_buy#v=onepage&q&f=false [Accessed 28 January 2014].





Word Count: 871

Level: EAP 5                  
Time: 18:00-19:30
Room: G.3



FINAL DRAFT

Advertisements are usually win-win methods for both companies to make their products visible for people around the world and customers to find what they are in need of. However, there are some situations when they become win-lose methods, supposedly when they are misleading. There are many methods to make an advertisement misleading. This essay will thoroughly describe how different methods works on each situation and give an example for each of them.

The first method is called “Guarantee without a Remedy Specified”. The technique is used when companies do not specify what they will make up to the customer, if the products are under expectation. When the product fails, unlike what the customers expected, the companies are free to do very little. However, the law says that the advertisers have to be clear on advertising, not only to determine an error but also to give a solution for the error (“False advertising,” 2014). For instance, the chill tonic advertisement (see Figure 1) below does not give any specific remedy because it only mentions, “No cure No pay”.



Fig. 1 A guaranteed-without-a-remedy-specified advertisement

The second method to make misleading advertisements is “No risk”. It is a strategy when the advertisers claim that there are no risks to try their products when clearly there are. For example, they may charge the customers’ credit cards for the products and they will offer a full refund if not satisfied. However, the risk of this kind of offer is quite enormous. The customers may not receive the products at all or the products can be something that the customers did not paid for. Sometime, they tell the company to permit a return but they are unable to do so (“False advertising,” 2014).

The third method of misleading advertisements is “Hidden Fees”. It is a confusion caused by products’ owners, who try to hide extract fees in small letters or use confusing terms (“False advertising,” 2014). It is an attractive method to advertise in order to gain the amount of the product sold. Obviously, it can lead to the misleading statistic when the prices that are shown on the products are not fully covered all of the amount of expense the customers will make, which means they may have to spend more for other services, such as taxes and step-by-step payments. As we can see from the picture (see Figure 2), which is an advertisement of a smart phone called iPhone. It shows the lowest price of the original price on the advertisement. However, this is a step-by-step payment method. This can confuse and attract consumers. Moreover, that price may exclude the taxes payment. This method can decrease the price shown on the advertisement.


Fig. 2 An iPhone advertisement

     The next principle method is “Bait-and-switch”. It is used when advertisers advertise an unavailable item to attract the clients to visit theirs shops. When the consumers arrive at the stores, they will be convinced to buyother similar products at higher price. “Bait-and-switch” is legal in some countries, especially in the United States (“False advertising,” 2014). For example, the products that are advertised on newspapers at special occasion such as New Year or Christmas sometime are not available at the stores. However, by the time the customers learn about the unavailability, it will be too late since they will be at the store already. So they will be forced to try other products.

     Similarly, there is another method to mislead advertisements. It is the “misuse of the word ‘free’”. “The usual meaning of ‘free’ is ‘devoid of cost or obligation’” (“False advertising,” 2014). Sellers use this word to give away products that the prices are already included in overall. The "buy one, get one free" sale is the most common example. The meaning of "free" of the second item is not normal, since, to get it, the consumers have to fully pay the fee of both items on the first item (“False advertising,” 2014). For instance, hotels advertisements such as “Stay two nights, get the third free” can be misleading because the price of the third night may be included in the first two night already.

Last but not least, comparative statistics can also be misleading advertisements. This method can be notified when advertisers use comparative words such as “better” or “more” in the advertisements. In this method, advertisers prove their products to be better but they usually left out what the products are better than.  When people read the advertisements with only half of the information was given, their brains will automatically create another half("Chapter 7:the semiattached figure,").  For example, this method was applied on a toothpaste brand called “Arm & Hammer” (see Figure 3). Clearly on the package, there were letter saying “3 shades whiter, clinically proven”. Normally, customers will think that the toothpaste will make their current teeth to become 3 shades whiter. However, this toothpaste might mean that it makes the teeth 3 shades whiter than not brushing or than using other products.


Fig. 3 A toothpaste package

All in all, advertisements are not trusted sources for people to believe and they are very dangerous for ignorance because they might get confused or misled by the advertisement. I believe that if there are no improvements in any time soon, the world will face an economic problem because people will stop believing in the advertisements and stop buying products. So further rules must be made to stop this type of lie and some other rules must be strengthen before it becomes a serious problem.






Bibliography

1.      False advertising. (n.d.). In Wikipedia. Retrieved February 25, 2014, from:http://en.wikipedia.org/wiki/False_advertising
5.      Chapter 7:the semiattached figure. (n.d.). Retrieved from: http://www.calvin.edu/academic/economics/faculty/bios/HaneyDocs/page-58978274.html




Unfair Comparison (4th draft)
            There are many forms of statistics that people use every day. Unfair comparison statistics are statistics that compare two unequal items or concepts. One data is always better than the other which is why it is misleading. Generally, this type of statistics is popular in the business world. Most companies utilize this statistics to bring out a congenial image for their products as well as to make the product sounds better than it really is. This statistic is also used in numerous government departments and some kind of reports in order to make up surprising outcome or how better one item is, compared to the other item. This type of statistics served as a beneficial tool for most companies or organizations. However, it can also cause serious issues for many people and companies. For instance, it can lead one company into an atrocious image which can possibly make a company go bankrupt or lose their businesses because being compared to a bigger or better company will make one company look worse than it really is. It is an extremely bad influence to societies because most people think that every information that comes out as a survey or a statistic is reliable and trustworthy. This is why a lot of confusion was created and it harms people’s health and property. Unfair comparison is technically right but if people think about it carefully, we will figure out that it is actually so misleading that we cannot believe in it. This type of statistic is very tricky; it can easily convince people to purchase something or to believe in something.
            Many examples of unfair comparison can be seen in statistics of every sort. One of these examples, used by crop fertilizing company, includes the comparison of crop yields on selected years (Beri,2010). This method is used in order to showcase how much improvement their products can achieve. It can be done by simply selecting the crop yield of the year when there are occurring disasters including drought or inadequate rainfall and compare it to the year when there is no such issues. False information, which states the differences between the crop production before and after the used of the fertilizing product, can then be added to convince the audience and show how much of an impact their products can make. This comparison will definitely give the product a better picture (Beri, 2010). Another example of unfair comparison occurs when there is a comparison of complaints between companies including airline business (Lori Alden, 2005-7).
















As shown in the figure above, it appears to viewers that the airline companies that got the most complaints which in this case is United Airlines, is the worst airline company; while the least complaints of airlines including the Alaska Airlines, Southwest Airlines and Continental Airlines are the best airline companies. This statistic is confusing since it does not provide any related information or reason why United Airlines got the most complaints. This could be due to numerous reasons including the amount of passengers on the United Airlines.
            Even though, the unfair comparison or comparing apples and oranges are complicated and misleading, there are many ways to make the statistics more accurate. Firstly, the writers can compare best information with best information (Lori Alden, 2005-7). Secondly, if the writers do not want to compare best information with best information or they did not get enough details about the best data consequently, they can compare lowest information and lowest information, according to Lori Alden (2005-7). It is another strategy to make the statistics more reliable (Lori Alden, 2005-7). Also Lori (2005-7), illustrates that if the writers want to make a statistic for their business such as their products, they can compare their products with another companies’ products by comparing all to all. So, all the details are necessary because everyone can see and examine it. Thus, it is better to compare all to all because some people might not want to see only best data with best data or lowest data with another lowest data but they want to know all the information both good and bad consequently, it is a good way to compare objects or data by using this strategy (Lori Alden, 2005-7). There are many strategies of making the accurate statistics and all of them are usually used in the statistics to make it more reliable because the unfair comparison or comparing apples and oranges might make people get the wrong information.



Bibliography:
-          Beri, G. C. (2010). Business statistics. (3rd ed.). New Delhi: Tata McGraw Hill Retrieved from http://books.google.com.kh/books?id=tWmoP49v1cIC&printsec=frontcover&source=gbs_ge_summary_r&cad=0

-          Lori, A. (n.d.). Statistics can be misleading. Retrieved from http://www.econoclass.com


 

             To sum up, there are various types of misleading statistics such as sampling bias, misleading graphs, misleading advertisements and unfair comparisons. People should be aware of all kinds of statistics as they can be misleading in many cases. Therefore, precautions should be taken in order to avoid misinterpreting the data. As can be predicted, in the future, if misleading statistics still exist, people will lose their belief in statistics and this will cause a substantial impact on everything ranging from bankruptcy of companies to an economic crisis.

No comments:

Post a Comment