Collection of data complete.pptx

Chapter 2: Collection of data “Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write”. ∑ ---H.G Wells Beretu. T. ([email protected]...

Chapter 2: Collection of data “Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write”. ∑ ---H.G Wells Beretu. T. ([email protected]) LERANING OUTCOMES Upon completion of this Chapter, you should be able to: distinguish between primary and secondary data sources examine various sources of primary data appreciate the art of questionnaire design distinguish between probability and non-probability samples conduct a sample distinguish between different methods of data collection. METHODS OF DATA WILL DEPEND ON: NATURE OF THE PROBLEM TIME AVAILABLE MONEY AVAILABLE DATA SOURCE DEGREE OF ACCURACY REQUIRED SOURCES OF DATA: Where to get them? PRIMARY DATA SECONDARY DATA New data Already from scratch existing data  Primary VS Secondary data Dimension Primary data Secondary data Definition First-hand data gathered Data already collected by by the researcher someone else earlier themselves Data Real time data Past data Cost Expensive Inexpensive Specific Specific to the needs of Might not be specific to the researcher the researchers needs Collection time Lengthy Short NB: Any data source (external or internal) used for a task other than that for which it was originally collected can be described as secondary data.  Secondary data External secondary data It is divided into internal and It is data that Tax records and social external secondary data. security data Census data (the Stats SA, health records, books, journals, social Internal secondary data media monitoring, internet searches, and other online data. Secondary data is not limited to that from a different organisation. It can also come from within an organisation itself such as sales reports, HR filings, payroll figures, emails and metadata  PRIMARY DATA An original data source collected specially for the purpose in mind Four main methods of data collection: 1. Face-to-face 2. Phone 3. Post 4. Via the internet PRIMARY DATA SOURCES You can obtain primary data by:  Investigating or experimenting  Observation  Focus groups  Conducting surveys using questions  1) Conducting an experiment Ask yourself: Are you doing anything to the people being studied? Deliberately impose some You are Practical treatment on manipulating a example: p17 individuals or variable objects  EXAMPLE If a researcher is interested in the eff ects of a new medication on headaches, the researcher would randomly divide a group of people EXPERIMENT with headaches into t wo groups. One of the groups, the experimental group, would receive the new medication being tested. The other group, the control group, would receive a dummy medication  Collecting data relies on watching or 2) Observations listening very carefully, and then counting or measuring events as they happen without any interaction with the individuals or Practical objects example: p17 Attitudes? Can these be Opinions? observed? Beliefs?  EXAMPLE Raeesa undertook took observation as a means of collecting data in the customer services call centre of a retail company. Her research focused on the training and quality assurance OBSERVATION of call centre staff. She observed the manner call centre staff dealt with complex customer issues, to understand how they used their discretion to deal with customers sensitively while seeking to adhere to their training and to any scripted parts of their telephone conservations with callers.  3) Focus Groups 0 1 Focus group Individuals are A small sample of respondents more willing to talk This may the target group is sometimes about things when facilitate the feed on each they can do so selected to learn design of other’s within a group how respondents comments discussion format questionnaires talk about the topic or other of interest. research tools.  EXAMPLE Evan Williams and Biz Stone conducted market research to fi nd out what tech users were looking for in social networks. They FOCUS GROUP recruited a small sample of the target population to explore and discuss a topic, such as what Facebook users were dissatisfi ed about. They discovered that the number one complaint was that the news feed was too cluttered. Using this feedback, the founders produced the idea behind Twitter – a simplifi ed social network for users to share news and opinions in 140 characters or less. 4) Survey by means of asking questions 0 02 1 Decide what questions will be Surveys are asked by a good way designing a Predetermi of questionnaire ned set of gathering a and choosing questions large how these that is amount of questions will be given to a data. put to the sample people. Some of the most used methods to collect data when conducting a survey using a well-designed questionnaire are:  Personal interview  Telephone interview  Mail questionnaire  Electronic questionnaires  Personal interviews Data is Can clarify difficult obtained questions and verbally show visual and face displays or to face. products Biggest advantage: detailed feedback. The interviewer must tell Obtain in-depth respondents responses from beforehand how respondents  Telephone interviews Presentation Travel time and costs are of the eliminated! questionnaire by telephone. Major disadvantage: people may not answer! Less costly than People are more personal interviews open in their and can be opinions as there conducted over is no face-to-face wider geographical contact  Mail questionnaires complete and return a questionnaire which they Disadvantages of this receive in the mail, in a method include a low newspaper etc. response rate, Major disadvantage: no allowance for illiteracy (4,4 millions of adults) etc. Can cover a large Respondents can sample since it is remain anonymous if relatively cheap to they desire and will administer therefore be more open and honest in  Electronic questionnaires Relatively new and fast-growing Disadvantages of this method method include a low response rate, Quick and inexpensive but often less detailed! e-mail containing a link is The least expensive sent to respondents; way to reach the they are asked to click on greatest number of the link to fill in a people – globally questionnaire  QUESTIONNAIRE DESIGN t r a t ive: i n is s, adm me, addres na A date, etc QUESTIONNAIRE i fi c a tion: s CAN BE DIVIDED clas ex, age, s race, s t atus, IN 3 SECTIONS: ta l mari io n, etc f p a t occu a tt er o su b ject-m (the ry inqui ns). io quest  QUESTIONNAIRE WORDING All questions should be appropriate to the research topic. Each question should be short and easy to understand. Questions should be unbiased Questions should not be phrased emotively. Place questions that may evoke an emotional response near the end of the questionnaire since they may influence responses that follow. Questions should not be offensive or embarrass the respondent. Wherever possible, a choice of answers should be given (closed questions). Make sure that every possible answer is covered. When this is not possible, adequate space should be given for answers. Confidentiality should be assured.  ADD TEXT 35% 1 TYPE OF 2 Frequency – 5 QUESTIONS point: 3 1 – Never YOUR2 – CONTENT Rarely 3 – Sometimes 4 – Often 5 – Always WE CAN ASK WRITE SOME TEXT HERE Closed Closed Open-ended Open-ended ADD TEXT questions questions questions questions  REPORT Based on all questionnaires Descriptive statistics Recommendati ons  Selecting a sample  Usually, Statistics South Africa interviews every household in South Africa to generate stats about the country.  In view of the foregoing, if companies do not have a certain budget to study the whole population, they would then select a reasonable number from the population so that the study can be able to make inferences.  The reasonable portion selected above would then form what is known as a sample.  ADVANATGES OF SAMPLING:  Reduces cost; collection time is reduced; overall accuracy is improved  SAMPLING LAWS: Law of statistical regularity: the selected sample must be a reasonable number of population.  Law of inertia of large numbers : large numbers show more stability than small ones Sample design  Sample design represent the way we select our sample from the population. For instance , we may decide to select our sample randomly or deliberately based on certain characteristics.  Types of sampling strategies (methods): Non-random/non-probability sampling methods  Convenience sample: the researcher chooses the readily available or willing to participants. It is convenient for the researcher to select the first few sample items.  Judgement sampling: the sample consist of items deliberately chosen from the population based on experience or judgement of the researcher.  Voluntary response sampling: this method selects those participant who volunteer to give responses.  Snowballing sampling: the sample is selected based on referrals from other survey respondents. For instance, if you interviewing late comers in the class, you might want to ask the current respondents whether he has referrals for you or not.  Continuation  Quota sampling is a non-probability sampling method that relies on the non-random selection of a predetermined number or proportion of units. You first divide the population into mutually exclusive subgroups (called strata) and then recruit sample units until you reach your quota.  E.g. A researcher wants to survey individuals about what laptop brand they prefer to use. He/she considers a sample size of 500 respondents. Also, he/she is only interested in surveying nine provinces in South Africa. Here’s how the researcher can divide the population by quotas:  Gender: 250 males and 250 females  Age: 100 respondents each between the ages of 16-20, 21-30, 31-40, 41-50, and 51+  Employment status: 350 employed and 150 unemployed people.  (Researchers apply further nested quotas. For eg, out of the 150 unemployed people, 100 must be students.)  Location: 50 responses per state  Sample design continued..  Simple random sampling: (simple) random sampling means that every participant has an equal chance to be selected in from population.  In a simple random sample, every member of the population has an equal chance of being selected. Your sampling frame should include the whole population.  To conduct this type of sampling, you can use tools like random number generators or other techniques that are based entirely on chance.  Example: Simple random sampling. You want to select a simple random sample of 20 stats students in stats class. You assign a number to every student in the class database from 1 to 20 and use a random number generator to select 20 numbers.  Sample design continued..  Stratified sampling: would be selecting a sample that has the same characteristics within a population.  In stratified sampling, researchers divide subjects into subgroups called strata based on characteristics that they share (e.g., race, gender, educational attainment).  Once divided, each subgroup is randomly sampled using another probability sampling method. Sample design continued..  Systematic sampling : assigning numbers to certain objects of the population and therefore, select numbers randomly.  Systematic sampling is like simple random sampling, but it is usually slightly easier to conduct.  Formular is: k=N/n. k= systematic sampling interval; {N}= population size and {n}= sample size  Every member of the population is listed with a number, but instead of randomly generating numbers, individuals are chosen at regular intervals.  Example: All stats students are listed in alphabetical order. From the first 10 numbers, you randomly select a starting point: number 6. From number 6 onwards, every 10th person on the list is selected (6, 16, 26, 36, and so on), and you end up with a sample of 100 people.  Sample design continued..  Cluster sampling: some population have a group which within themselves represents all of the views of general populations.  Cluster sampling is a method of probability sampling that is often used to study large populations, particularly those that are widely geographically dispersed. RANDOM/PROBABILITY SAMPLING METHODS  Summary  Research process : solving a problem where the researcher ask question that he or she wishes to answer.  The purpose of statistics: to collect and analyse information using statistical measurements.  Methods of data collection depends on the questions; problems and the type of sources.  Primary and secondary sources of information.  Questionnaire design: ensure that questions are in conjunction with the research objectives.  Types of questions to ask in a questionnaire: open and ended questions.  Sampling design : the way a sample is selected. Either randomly or non-randomly.