Questions on Understanding Data
Questions on Understanding Data
13-3 What is the difference between attribute data and variable data? Give several examples of each.
Attribute is the characteristic of a product. Attribute data is the data derived from the quality of the product, rather than quantifiable numbers. This data is in either yes or no; go or no go format. Though this data can be sometimes classified as subjective, it is compared to similar data to make it countable or discrete. Variable data is numerical information or the measured value of a product. This data can be quantified according to size, length, breadth, temperature.
First examples of attribute data is, how many times has the machine failed to perform certain task, for the same example the variable data will be, the error the machine recorded each time it failed. Second example of attribute data will be ranking of products from best to worst while the variable data will be the variance of the product from the mean. Third example of attribute data can be number of patients coming to a clinic with similar range of fever. The variable data for this will be the record of temperature ranges.
13-5 Suppose an airline had over 6000 missing pieces of luggage last year, and this number represents 0.2% of the total pieces of luggage handled. What types of data would you request from the airline to help you understand the nature of the problem and how to solve it?
- which airports (source and destination) records maximum missing luggage – after knowing this the airport authority can be warned and put extra effort to fix the problem
- What is the nature of all missing luggage (were the luggage send late after check in, were they in transit) – If the transit luggage are missing, they can be fixed, if only late luggage are missing, either travellers or ground staff need to be more alert
- Which alliance airline is having maximum issues – Many a times airline use their alliances with other airlines. This way we can check if any alliance is causing problems.
- What type of luggage is generally missing (bag, suitcase, others) – With this information an extra care can be taken for those specific pieces of luggage.
13-6 Suppose a bicycle manufacturer is experiencing a high rate of returned merchandise from its customers. The most frequently cited reason for the return (95%) is “difficulties with assembly.” What types of data would you request from the bicycle manufacturing company to help you understand the nature of the problem and how to solve it? Be sure to state why you believe the data you request would help you, and state how it would help you.
- Which part is recoding maximum difficulty in assembling? Part identification is a must, whether it is handle, seat, brakes or anything else.
- Which supplier is supplying those parts? If a part has problems we need to check is a particular supplier is causing problem.
- Which assembly unit is experiencing the maximum difficulty: Is this happening at a particular assembly point? Does the assembly unit have skilled manpower?
- Transportation issues: Is any particular location having an issue? Is the issue with related to transportation? Are some damages happening during transit?
13-7 What is the purpose of sampling?
While examining a bigger population it is impractical to consider the whole population for studying a trait. Thus, sampling is done to study the characteristics of the total population or product using a small sample. The sample is examined and the results are applied for the whole population.
13-8 What is sampling error?
Samples do not exactly give a true picture of entire population. Thus sampling error occurs. This error is the difference between the sample characteristics in comparison to the actual population.
13-10 What factors should be considered when choosing a sampling scheme?
Factors considered while choosing a sampling scheme are as follow
- Quality and accuracy of data of the sampling scheme – This is most important. More reliable data, more accurate will be the results
- Extra cost and people needed to carry out the sampling scheme – Sampling is an intensive job and is done with large volumes of data. Extra cost as well as trained professionals are required to carry out sampling.
- The timelines within which the data is taken. This is because samples keep changing with time. A relevant sample scheme can be accurate only if sampling is done within strict timelines.