Six Sigma Data Classification, Sampling, and Collection

Match each data coding method to the example of its use. Adding or subtracting a constant Coding by substitution Coding by truncation Multiplying or dividing by a factor

Adding or subtracting a constant - A data frequency of (99, 199, 299, 399, 499) is coded as (100, 200, 300, 400, 500). Coding by substitution - Collectors enter the number of 1/8-inch intervals away from the mean of 10, rather than entering the measurement itself. Coding by truncation - 553001, 553002, and 553010 are entered as 01, 02, and 10. Multiplying or dividing by a factor - A frequency distribution of (110, 210, 300, 410, 520) is represented as (1.1, 2.1, 3.0, 4.1, 5.2).

Match the types of data collection to their characteristics. Each type may match to more than one characteristic. Automated Data Collection Manual Data Collection

Automated Data Collection - Free from human error. May require expensive equipment. Provides objective information. Is accessible by multiple sources. Manual Data Collection - Typically subject to bias. Uses very simple tools. More time consuming.

Match the types of data to their examples. Each type will match to more than one example. Continuous Discrete

Continuous - The range of temperatures that are comfortable in a restaurant. The time it takes to serve coffee. The weight of each plate in a crockery set. The monetary cost of recovering from bad publicity. The temperatures a dish can withstand in the oven. Discrete - The number of times customers leave positive feedback. The people who have access to sensitive company information. The small, medium, and large sizes available for a cardigan. The number of customers who qualify for a discounted price.

You're using automated data collection, and need to distinguish between different types of information technology. Match technologies to their characteristics. Technologies may be used more than once. Data warehousing Data mining Online analytical processing (OLAP)

Data warehousing - Data is collected as metadata from the source systems. Metadata is used to provide operational feedback. Data mining - It uses a relational database to discover patterns in large datasets. It transforms extracted data into information for process improvement. The end users receive data from departmental warehouses. Online analytical processing (OLAP) - It uses a relational database to analyze data from multiple perspectives.

You want to find out which T-shirt size sells the most. What type of data will you collect?

Discrete Data.

Which principles should you follow when collecting data samples?

Ensure that the sample represents the entire population you're studying. Collect a sample that's big enough to detect trends. Ensure that the sample is free from bias. Minimize heterogeneity in the sample.

Which questions can be answered by collecting continuous data?

For how many hours was each machine idle? What were the factory temperatures over the past month? What's the weight of the AA batteries produced?

You are a Six Sigma Green Belt in a Customer Service Department. Match the types of check sheets with examples of when you'd use them. Frequency plot check sheet Basic check sheet Traveler check sheet Location check sheet Confirmation check sheet

Frequency plot check sheet - Document which days calls are dropped during the customer complaint resolution process and how many times it occurs on each specific day. Basic check sheet - Find out how many of each type of mistake occurs when resolving a customer complaint. Traveler check sheet - Track the stage duration and number of mistakes it contains for each stage in the customer complaint resolution process. Location check sheet - Identify the parts of a customer complaint resolution form that contain mistakes by adding marks to a copy of the form. Confirmation check sheet - Ensure that each step in the customer complaint resolution process is being followed.

Match the levels of measurement to their characteristics. Interval Ratio Ordinal Nominal

Interval - The distance between ranks is equal and there is no absolute zero. Ratio - The variables have a fixed zero point and can be compared using percentages. Ordinal - Variables are ranked relative to each other, and the distance between ranks isn't indicated. Nominal - Variables are discrete but not placed in a particular order.

You are conducting a survey where people need to identify their original hair colors.What type of measurement scale is used?

Nominal

In a survey of people's level of trust for certain websites, participants rank the sites by checking one of three options: no trust, some trust, or high trust. What type of measurement scale is used?

Ordinal

Match the data classifications to their examples. Each classification may match to more than one example. Qualitative Quantitative

Qualitative - The dinner plate is rectangular. The computer monitor has a white frame. The coffee is served in a ceramic mug. The toothbrush has a comfortable handle. The bowl can be microwaved. Quantitative - The laptop weighs two pounds. The soup takes 45 minutes to prepare. The cup of coffee has one inch of foam at the top. The computer monitor has a resolution of 1920 by 1080 pixels.

You want to find the standard deviation for T-shirt dimensions within particular production lines, as well as between the production lines. What's the best sampling method to use?

Rational sampling.

Match sampling method characteristics to the methods they're associated with. Some characteristics may match to more than one method, and some methods may have more than one match. Stratified sampling Systematic sampling Rational sampling Simple random sampling

Stratified sampling - The population is divided into groups. The size of each group in the sample is proportional to its size in the population. The population groups should be proportionately reflected in the sample. The population is diverse. The population is heterogeneous. Systematic sampling - Every "nth" unit is sampled. It's typically used to collect real-time data from process operations. Rational sampling - The population is divided into groups. It's typically used to collect real-time data from process operations. It's helpful for calculating estimates for standard deviation. It can be used as the basis for effective control charts. Simple random sampling - Each unit has an equal chance of selection. You have an accurate database of the population and no interviews are required.

What are some considerations to make when developing your data collection plan?

The data sources for collecting the best data. That the right data is being collected. The tools and teams involved in collecting the data. The purpose and goals of the data collection. The people involved in the data collection effort. The format the data should be collected in. The best place in the process to collect data from. Reasons for collecting the data and what will be accomplished by collecting the data.

When should you use stratified sampling?

The groups are more homogeneous than the main population. The population has groups that should be proportionately reflected in the sample.

When should you use simple random sampling?

The population is homogenous. No interviews are required. You have an accurate database of the population. The process is not well understood

Which best practices should you follow when collecting data?

Use an appropriate error correction system for digitally stored data. Record the before and after values for variables that change over time. Provide people with complete training on the process. Record the order in which data is collected. Maintain accurate data collection tools. Record classification identification information during data collection. Avoid unnecessary data rounding. Avoid discounting outliers. Regularly screen the data for entry errors.

Which questions can be addressed by collecting discrete data?

Which screen has the best balance between strength and weight? Does process A, B, or C yield the best quality product? How many products are defective per 10,000 produced?