Thursday, March 12, 2009

March 14, 2009 - Overview of Statistics I

A. Overview of Statistics

Statistics is defined as the science of collecting, organizing, analyzing, and interpreting data in
order to make decisions. It uses mathematical formulas for analysis, and it involves the
understanding and interpretation of the results.
Statistics involves studying two types of data sets—population and sample.

Population:
• It is a collection of all outcomes, responses, measurements, and counts that are of interest.
• The numerical description of a characteristic of a population is called a parameter.
• A count or measure of an entire population is called a census.
Sample:
• A subset of a population is called a sample.
• The numerical description of a characteristic of a sample is called a statistic.
• A count or measure of a part of a population is called a sample.

The study of statistics is divided into two major branches:

Descriptive statistics: The branch of statistics that involves organization, summarization, and
display of data

Inferential statistics: The branch of statistics that involves the use of a sample to draw
conclusions about a population

B. Data Classification

A clear understanding of the meaning of the term “data” is central to the study of statistics. Data
consists of information related to observations, counts, measurements, or responses.

Class Activity

Qualitative: It consists of attributes, labels, or nonnumeric entries. For example, gender—male
or female—refers to qualitative data.

Quantitative: It consists of numerical measurements or counts. For example, age—1, 2, 10, or
20—refers to quantitative data.

Another way to classify data is by its level of measurement, which determines the relevance of
statistical calculations. Using this categorization, we have nominal, ordinal, interval, and ratio
data.

C. Experimental Design

In the real world, statistical results can mislead or misrepresent the facts if the research
conducted does not use proper procedures. Therefore, while making a decision based on
statistical analysis, we should be aware of the process used to obtain the data and the potential
misuse of the data. Given here are some guidelines for designing a statistical study:

1. Identify the variables of interest and the population of the study.
2. Develop a detailed plan for collecting data.
3. Collect the data using any of the following methods:
i. Doing an observational study
ii. Performing an experiment
iii. Using a simulation
iv. Using a survey
4. Describe the data using appropriate descriptive statistics or graphs or both.
5. Interpret the data and make decisions about the population by using inferential statistics.
6. Identify any possible errors.

Surveys can be done by taking a census or using a sample. To ensure an accurate representation
of the population, appropriate sampling techniques should be used; otherwise, the results from
the study may be considered invalid. Some commonly used sampling techniques are random
sampling, stratified sampling, cluster sampling, systematic sampling, and convenience sampling.

HOMEWORK: DESIGN YOUR STUDY & POST FOR EACH DESIGN NUMBER

10 comments:

  1. Tonya Cox


    1. The size of Canyon County and how many Canyon County residents live outside the city limits.
    2. I would go to the court house in Canyon County and look at the information they might have.
    3. I could do a survey by going to say a gas station close to the city limit line and over an 8 hour period, I could ask customers who come in, “Do you live in Canyon County? If so, do you live in city limits or outside city limits?
    6. There could be errors in the sense that maybe more than one person might get asked more than once. But hopefully that person would let you know that you had already asked them that question, so that would help the problem.

    ReplyDelete
  2. 1) Ironically, I think I would be interested in using the Call of Duty idea, more specifically Call of Duty World at War. I think I would focus on the XBOX Live population inparticular. How many XBOX Live patroons from the United States versus the foreign players, on specific times of day and day of the week.
    2) I would simply play on XBOX and ask some questions here and there, and collect my data by speaking to fellow players online. I would also check with XBOX Live directly and see if there is any data.
    3) I would probably utilize the observational and survey techniques. I beleive these methods would render sufficent information.
    4) To be continued....
    5) To be continued....
    6) The only errors I can think of at the moment would be the occasionaly liar, and the lack of trusting a complete stranger. The information may be difficult to gather, it could take more time than I would be willing to put forth. So, basically it boils down to time management and gather sufficent data in a reasonable span of time.

    ReplyDelete
  3. Mel Bailey

    1. I think it would be interesting to know how many people know the name of the anti virus software on their computers. I am not limiting the population of study to any age group!
    2. I am a cosmetologist so I will ask people as I am cutting their hair.
    3. The data I collect will be an experiment/survey.
    4.
    5.
    6. A possible error might be not getting very many participants if work is slow.

    ReplyDelete
  4. John Drake
    1. I will determine the most popular drink served at Terapin Staion on a Sunday night during Industria.
    2. data collection is a simple matter of haveing the bartender record each drink as he serves it.
    3. This colection can be done over several consecutive Sundays to provide the widest sample, and then averaged by calculation.
    4.
    5.
    6. Errors may ocurr if the bartender forgets to record or if i skew the data by ordering drinks myself.

    ReplyDelete
  5. 1. I am interested in finding out if internet users spend more time reading news and entertainment to keep up on current events or using email and IM to stay in touch with friends.
    2. I will use online survey sites like ask500 to collect the data.
    3. I will use an online survey site.
    4. I will use graphs and a summation of findings along with possible identified errors in the data.
    5. I will use inferential statistics to interpret the data and postulate on the level of expertise of using technology.
    6. Possible errors include the limited population of users of that one survy site, which may or may not be a good representation of the broader internet using population.

    ReplyDelete
  6. Statistical Study on World of Warcraft Server population.

    We want to get a good picture of the total population of the United States Alliance characters on the server named, Naz'Grel.

    1.The area of interests are (Total population in United States, Alliance, Characters, Server named Naz'Grel).

    2.Several web sites contain statistics on each server as well as subscription details. Once I have such data, I can use it to populate the missing elements in the equation.

    3.In order to collect this data, we will perform an observational study or research on the various web sites.

    4.According to several websites, there seems to be a common number of subscribers which adds up to about 12 million since the start of 2009. There are a total of 243 servers that contain both Horde and Alliance characters. There is never a perfect scenario of the ratio of 1:1 on any server. According to Warcraftrealms, the number of alliance characters on this server is roughbly 16,383. The number of horde characters are about 3,827.

    The image below shows a tiny pie graph; if you're interested, paste the address in your URL

    http://www.spectabulous.com/files/school/eg381/nazgrel.jpg

    According to the pie chart, the blue field represents alliance characters, and the red field represents horde.

    5.If we add the total population together from each side, we conclude with 20,210 characters. Converting that to a percentage, we can conclude that 81% of the population is alliance, and 19% are horde. If we turn this into a ratio, we can conclude that there are about 4.27 alliance to 1 horde (not exact, but close). If we wanted to take things further, we could add the variant of 243 servers. With that in mind, we can multiply 20,210 by the 243 servers available, we would get a total of 4,911,030 characters which doesn't make sense considering that there are 12 million subscribers (give or take). This concludes that there are heavier populated servers than Nazgrel.

    6.There is always room for error in any heavily statistical assessment, especially in this case since people may have created, deleted, or have gotten their accounts banned after the data was collected. It is also possible that some accounts may have moved to a different server.

    DISCLAIMER: These values are based on figures collected on various web sites that may not be affiliated with World of Warcraft.

    World of Warcraft key to Vivendi's 2009 revenues. Retrieved March 20, 2009, from Joystiq Web site: http://www.joystiq.com/2009/03/02/world-of-warcraft-key-to-vivendis-2009-revenues

    Realm Status. Retrieved March 20, 2009, from World of Warcraft Web site: http://www.worldofwarcraft.com/realmstatus/compat.html

    Graphical Display Census. Retrieved March 20, 2009, from WarcraftRealms Web site: http://www.warcraftrealms.com/census.php?serverid=742&factionid=2&minlevel=1&maxlevel=80&servertypeid=1

    ReplyDelete
  7. Study of Lost Jobs:
    Evan D.

    1. People in or around the Boise area who have lost jobs in the last 6 months due to the economy.
    2. A simple survey would collect the first portion of data, then just compiling it. Most likely send survey by mail.
    3. iv. using a survey
    4. Bar graphs would be most useful
    5.
    6. People who may believe they've lost their job because of economy, but have not could skew results harshly. Jobs in sensitive areas of the market, where layoffs are common.

    ReplyDelete
  8. 1. I would think it interesting to see how different populations migrate into the USA. As in, what types of visas (fiancée, work, asylum, refugee) and what stage of their lives (young, old, married, single) as well as from which countries.
    2. The data, for the most part should be readily available through census data and other surveys conducted by the government
    3. TBD
    4. TBD
    5. TBD
    6. There could be many errors that stem from data collection to interpretation. Making sure that all data is relevant and correct will be first priority.

    ReplyDelete
  9. 1. Identify the variables of interest and the population of the study.

    - Boise State Broncos Fans
    - City of Boise

    2. Develop a detailed plan for collecting data.

    Sales of Broncos tickets, both individual and seasonal.
    Test decibel samples on or around Boise State during games.
    Has advertising dollars increased on local television stations?

    3. Collect the data using any of the following methods:

    Doing an observational study.

    4. Describe the data using appropriate descriptive statistics or graphs or both.

    I would describe to data using BSU Broncos Pennants.

    ReplyDelete
  10. It may be interesting to find out how many people rent DVD movies from movie rental stores/redbox/Netflix. Specifically DVDs and NOT Streaming like Netflix provides via PC, streaming boxes, or Xbox 360. There is much change in the ways people watch movies today. Go pay big bucks to watch a movie in a theater. Wait until the movie is released on DVD to watch it at home. Different kinds of quality video in High Definition HD movies, or blu-rey movies. One could wait for the providers, “Netflix” in order to stream movies over an Internet connection. I want to know about the Discs. There are costs involved in using a disc. Must have a DVD player. Must rent or purchase a dvd disc movie. Must have electricity and a television/projector to watch it on. What are the costs of the other ways to watch a DVD movie like streaming, or going to a theater. Changes and future changes or trends of DVD viewing.
    1. Could ask people at a DVD rental store. Could ask students at a school.
    2. Survey people around the area “Meridian” and ask about the number of DVDs they purchase or rent in a month.
    3. A survey could work. In the past month how many DVDs have you rented or purchased?
    4. Graphs could show great results.
    5. What is the probability that a person in Meridian will rent/buy a DVD in the month sampled? Purchase a DVD perhaps in future months?
    6. Times of the day spent surveying at a Store. Times that the store are open to public.

    ReplyDelete