GCNU310A Notes on Statistical Surveys PDF

**Unit Title** **Discuss various matters and issues related to conducting a statistical survey** **Unit Code** **GCNU310A** **1 Data Sources** **Primary Data:** Data that are collected at first hand for the purpose of a particular statistical (統計的) enquiry (調查). E.g. Census (人口普查) of...

**Unit Title** **Discuss various matters and issues related to conducting a statistical survey** **Unit Code** **GCNU310A** **1 Data Sources** **Primary Data:** Data that are collected at first hand for the purpose of a particular statistical (統計的) enquiry (調查). E.g. Census (人口普查) of population by government. **Secondary Data:** Data that are being used for some purpose other than\ that for which they were originally collected. E.g. A company uses published statistics from\ government to estimate his business growth. **2 Advantages and Disadvantages of using secondary data** +-----------------------------------+-----------------------------------+ | **Advantages** | **Disadvantages** | +===================================+===================================+ | - **Quick, convenient (方便的) | - Collected for a different | | and less costly** | purpose, may violate (違犯) | | | personal privacy (私隱) | | - Different sources:\ | ordinance | | Using secondary data can give | | | us access to | - **Collected for a different | | otherwise-unavailable sources | purpose, may not fit the | | | purpose of survey** | | - **Ready to use** | | | | - **Information out dated** | | - Preview:\ | | | Secondary data allows the | - Accuracy (準確): unknown | | researcher to extend the | sources of error/bias (偏見) | | \'time base\' of their study | | | by providing data about the | - Limitations (限制) of | | earlier state of the system | representativeness (代表) and | | being studied. | validity(正確) | | | | | | - Limitations of documents | +-----------------------------------+-----------------------------------+ **3 Examples of Data Sources** +-----------------------------------+-----------------------------------+ | Primary Sources | Secondary Sources | +===================================+===================================+ | - **questionnaires (問卷)** | - **Published Statistics | | | (統計出版物): Official | | - **interviews (面試)** | statistics (官方統計), | | | digests & abstracts(摘要)** | | - focus group interviews | | | | - Electronic (電子) Sources: | | - **observation (觀察)** | e.g. from search engines | | | (搜尋器) | | - case-studies (個案研究) | | | | - Documents: data supplied by a | | - portfolios | marketing | | | | | - experiments(實驗) | | +-----------------------------------+-----------------------------------+ **4 Data Collection Methods** ❄ Direct Observation Examples - Suppose you wish to know how people behave (行為) - counting the number of cars passing a tunnel (隧道) - **Actual (實際的) situation (情況):\ the actual actions or habits (習慣) of person are observed** - No disturbance (打擾): the method is applicable when it is undesirable (討厭的) for people to know an experiment is taking place; - Skill required: result of observation depend on the skill of the observer; - **No opinion: opinions and attitudes cannot be obtained by observation;** - Lengthy period: some actions cannot be obtained by \'one-time\' observation, but by a continuous and lengthy period of observation; - Expensive: it would be expensive to tie up (繫住) personnel in such tasks. ❄ Interview Interviews are conducted by asking the interviewee a series of questions. It is a widely used survey technique in social surveys. They can be conducted at work, at home, in the street or in a shopping centre, or some other agreed place. ❄ Personal Interview - - - - - - - - - **Disadvantages:** - Need to arrange interviews beforehand (預先) - - - Geographic limitations (地理的限制) - - - - - ❄ Telephone Interview - **Relatively (相對地) cheap** - **Quick** - - - - - - - **Disadvantages:** - - - - - - - **Easy to be refused (拒絕)** ❄ Experiments  Suppose that a company intends (打算) to introduce **a fire fighting equipment using a new technique, they are unable to speak to anyone who has used it in the past**. It may be that the only way they can gather data will be to carry out some experiments using the new fire fighting equipment. A set of experiments could be **designed and carried out with the results being carefully recorded.** **5 Questionnaires** Questionnaires are a popular means of collecting data, but are difficult to design and often require many rewrites before an acceptable (可以接受的) questionnaire is produced. - - - - - - No prior (優先的) arrangements are needed. - - - - - - - - - - - **No control over who completes it.** - - **Example 1** Suggest the best data collecting method for the following situations/characteristics: I. Interview II. Experiment III. Observation IV. Questionnaire V. Statistical publication / Internet +-----------------------+-----------------------+-----------------------+ | **Situations/Characte | **Data collection | | | ristics** | method** | | +=======================+=======================+=======================+ | a. | Average time of | I | | | traveling to office | | +-----------------------+-----------------------+-----------------------+ | b. | Number of typhoons | | | | affecting Hong Kong | | | | each year. | | +-----------------------+-----------------------+-----------------------+ | c. | Study time that | | | | students spend in a | | | | week. | | +-----------------------+-----------------------+-----------------------+ | d. | Number of visitors to | | | | a restaurant in a | | | | day. | | +-----------------------+-----------------------+-----------------------+ | e. | Lifetimes of | | | | different brands of | | | | batteries. | | +-----------------------+-----------------------+-----------------------+ | f. | Relatively cheap | | +-----------------------+-----------------------+-----------------------+ | g. | High response rate | | +-----------------------+-----------------------+-----------------------+ | h. | Cannot obtain | | | | opinions | | +-----------------------+-----------------------+-----------------------+ | i. | Accurate and in-depth | | | | information can be | | | | collected | | +-----------------------+-----------------------+-----------------------+ | j. | Collected data about | | | | a "new" | | | | equipment/techniques | | +-----------------------+-----------------------+-----------------------+ **6 Questionnaire Design** A questionnaire is a powerful evaluation tool and a popular means of collecting data. There are some basic rules should be noted in your design. ❄ Objectives (目標) of the Survey +-----------------------------------+-----------------------------------+ | The questions should follow quite | | | naturally from the objectives. | | | | | | When you design your questions, | | | always ask | | | | | | - What are the objectives of | | | the questionnaire? | | | | | | - How you intend to use the | | | information? | | | | | | - What do you want to know? | | | | | | - How would you answer it? | | +-----------------------------------+-----------------------------------+ **\ ** **7 Features of a good questionnaire** The qualities of a good question are as follows: - **Avoid long questions**. - **Avoid questions that are actually asking two questions.** - +-----------------+-----------------+-----------------+-----------------+ | Ambiguous | Clear Question | | | | Question | leading to | | | | leading to | Clear Answer | | | | Ambiguous | | | | | Answer | | | | +=================+=================+=================+=================+ | | How many number | | | | | of times did | | | | | you visit the | | | | | cinema the | | | | | previous four | | | | | weeks? | | | +-----------------+-----------------+-----------------+-----------------+ | A. B. C. D. | | A. 0 time | | | | | | | | | | B. 1 time | | | | | | | | | | C. 2 times | | | | | | | | | | D. 3 or more | | | | | times | | +-----------------+-----------------+-----------------+-----------------+ - Avoid questions with predicted (預料) response - Avoid double negatives questions to reduce confusion (混亂) - - - - - - Do not use emotionally (情緒上) loaded or vaguely (含糊地) defined words. Quantifying adjectives (e.g., most, least, majority) mean different things to different people. - Do not use unfamiliar (不熟悉的) words or abbreviations , for example **\ ** **8 Questionnaire layout** - **Good readability (易讀)** - **Good and clear instructions (指示)** - - - **\ ** **9 Population (總體) and Sample (樣本)** **Population:** All subjects possessing (具有) a common characteristic (特性) that is being studied. **Sample:** A subgroup or subset of the population. The population includes **all objects of interest** whereas the sample is **only a portion (一部分) of the population**. **Example 2 The chairman of a Youth Centre wants to find the average time the members spend in the Youth Centre daily. What is the population of the survey?** **Solution The population is [all the members of the Youth Centre]** **10 Sampling Methods (抽樣)** In order to get the reliable characteristics of a population by investigating a sample, the size of a sample must not be too small and the sampling method must be appropriate. The sampling methods can be classified into two main categories: **Probability Sampling (概率抽樣) and Non-Probability Sampling (非概率抽樣).** ❄ Probability sampling Any method of selecting a sample such that each sampling unit has **a specific and usually equal probability** of being chosen. E.g. Simple random sampling (簡單隨機抽樣), stratified sampling (層狀(分層)隨機抽樣) and\ systematic sampling (系統隨機抽樣). ❄ Non-probability sampling When we can ignore the probability of selection and choose a sample based on some other criterion, such as accessibility or voluntary participation(自願參與), we call it non-probability sampling. E.g. **Judgment** sampling (判斷抽樣)and **quota** sampling(定額抽樣). **Disadvantages of non-probability sampling:** - cannot be generated by probability methods - with **unreliable** factors, such as the researcher\'s experience, or even **on luck** - whether the sample was representative of the population - **cannot determine the estimates** - regarded as a \'last resort\'(最後一招) **\ ** **11 Probability sampling methods** Probability sampling concerns about the probability of each sample being chosen. One can define the probability of a sample being chosen. **❄ Simple Random Sampling簡單隨機抽樣** Sampling in which the data is collected using **random numbers** or lottery. **(抽籤)**. Telephone directory can be a source for random sampling to start with. **Advantages:** - Sample is selected with equal chance. - It is free from bias. - It is easy to be understood and used. **Disadvantage:** - Sample is not evenly distributed. **Example 3** A manager wants to select 5 staff from a group of 25 staff. Use the Random Number Table below to do it. Column ----- -------- --- --- --- --- --- --- --- --- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- --- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Row 1 9 2 7 1 1 0 8 1 0 2 7 5 9 9 1 0 3 8 6 6 2 9 4 8 7 3 9 0 9 3 5 2 1 6 8 4 1 0 8 1 9 3 2 4 9 2 1 7 8 6 1 0 1 6 6 3 5 9 8 6 8 9 4 7 7 6 4 0 9 7 6 3 6 3 7 3 9 7 9 3 3 7 9 1. 2. 3. 4. **\ ** **Example 4** The Student Society organized a Christmas party and the tickets numbered **101 to 250** were sold. Prizes will be given to **3 winners** in a lucky draw. Use the Random Number Table above to do it, starting from **row 1 column 1**. Column ----- -------- --- --- --- --- --- --- --- --- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- --- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Row 1 9 2 7 1 1 0 8 1 0 2 7 5 9 9 1 0 3 8 6 6 2 9 4 8 7 3 9 0 9 3 5 2 1 6 8 4 1 0 8 1 9 3 2 4 9 2 1 7 8 6 1 0 1 6 6 3 5 9 8 6 8 9 4 7 7 6 4 0 9 7 6 3 6 3 7 3 9 7 9 3 3 7 9 **Solution** 1. To determine the number of digit we want. Since the samples are selected from **the population** 2. Starting from row 1 column 1, the listed numbers are: 3. Hence, the ticket numbers of the 3 prize winners are\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\...\.... **❄ Systematic Sampling 系統隨機抽樣** Data is obtained by selecting every *k*th object.\ E.g. a number of households is sampled and\ their family income being studied. However, if we were sampling **paper waste** produced by households, and we decided to sample 100 households **every 7th day** (say every **Monday**). Our sample might not be representative, because Monday\'s rubbish would very likely **include the Sunday\'s** rubbish. Thus, the amount of waste would be biased and over-estimated. The bias would be avoided if we take samples every 5^th^ or 8^th^ day. **Advantage:** - Sample is **spread more evenly over the entire population** **Disadvantage:** - There may possibly present of hidden (隱藏的) periodic pattern. **❄ Stratified Sampling 層狀隨機抽樣** Sampling in which the population is **divided into groups (called strata) according to some characteristics**. For instance, the population might be separated into males and females. A sample is taken from each of these strata(層)using either random or systematic sampling. **Advantages:** - Stratified sampling can **get rid (擺脫) of bias in sampling**. - If the strata(層) are properly designed, they are more accurately reflecting characteristics of the population from which they chosen than other kind of sampling do. **Disadvantage:** - **The cost is high and time consuming** **Example 5** Stratified random sampling is used in a company to select a **sample of 50** staff. If the ratio of male and female in the company is 6 : 4, how many male staff should be selected? **Solution** Number of male staff **Example 6** In an IVE campus, there are **800** students in IT department, **1200** students in BA department and **2000** students in ENG department. Stratified random sampling is used to select a sample of **20** students for focus group interview. How many students in each department should be selected? **Solution** Number of student in IT [\$= \\frac{}{\\ + \\ + \\ } \\times \\ = \\ \$]{.math.inline} Number of student in BA [\$= \\frac{}{\\ + \\ + \\ } \\times \\ = \\ \$]{.math.inline} Number of student in ENG[\$= \\frac{}{\\ + \\ + \\ } \\times \\ = \\ \$]{.math.inline} **Example 7** Identify the following sampling method. **Situations** **Probability Sampling** ---------------- ----------------------------------------------------------------------------------------------------- ---- (a) In a factory, the 48^th^ battery of every 100 batteries produced is selected for a quality check. II (b) In a school, 20 out of the 400 boys and 30 out of 600 girls are randomly selected for a survey. (c) 50 students are randomly selected and asked about their means of transport to school. (d) In a shopping mall, a customer is randomly selected every hour to comment on the services provided. **12 Non-Probability Sampling** Non-probability samplings are used primarily as **a matter of convenience (方便)**, but there is **no valid** (有根據的) way of determining of the resulting estimates. **❄ Judgement Sampling** Personal judgement plays a significant role in the selection. One important use of such sampling is in testing markets for new products, such **test cities** are usually **not selected at random**; instead they are carefully chosen because of someone\'s judgement. - **Advantage:** **required sample can be selected with professional judgement** - **Disadvantage:** can be biased due to one's ignorance (無知) or negligence (疏忽). **❄ Quota Sampling** Interviewers are simply given quotas to be filled. - **Advantage:** can meet certain **preset requirement**(s) (預設要求). - **Disadvantage:** once the quota is set, interviewers are **granted flexibility** in the choice of sample members. Interviewers naturally tend to select individuals who are **most readily available**. 1. Data collection methods -- ProProfs Quiz 2. Types of Survey Sampling Techniques