Better Data Collection and Questioning Practices

Gathering information on process performance and waste often involves asking questions. How well the questions fit the situation or target the issue at hand, however, greatly affects the quality of data gathered for a lean Six Sigma project.

Often data collection efforts can utilize historical data. There are instances where this is either impossible, too time consuming or too costly to gather and evaluation teams must collect data manually. In order generate the type and quality of data needed for lean Six Sigma analysis; keep the following guidelines in mind:

  • Share the manner in which data will be used with everyone gathering and providing data. This provides a frame of reference for clearer communication and removes a source of possible bias by calming fears and concerns.
  • Include an objective third party in data collection, if possible.
  • Confirm that questions target the process or issue under evaluation and are not too general.
  • Create data collection plans and forms that take stratification factors into consideration. In evaluating processes for lean manufacturing Six Sigma improvement, for example, issues such as operator, machine, shift and suppliers must be noted. By covering these factors up front, a duplicate round of data collection is avoided and time spent in the evaluation process is not wasted.
  • Design data collection forms as simply as possible. Overly complex forms can skew results or generate inaccuracies.
  • Use a column format for collection forms. It is more easily translated into spreadsheet format, making the data available for lean Six Sigma analysis more quickly with less chance of complication or data corruption.
  • Invest an appropriate amount of time and effort in training the data collection team. Training can be as simple or detailed as warranted by the process under study and the team's level of expertise. Guidelines on the collection form, cheat sheets, one-on-one or hands-on training are all options worth consideration.
  • Confirm the data collection form includes space to log the collector's information (name, date, time of collection, etc). This information can be significant if questions arise in regards to the data or collection method.
  • Perform trial data collections before launching the real thing. This provides opportunities to check the quality of questions asked and adjust the scope if necessary. Quite often it becomes apparent that further refinement of questioning is warranted. Missing information, questioned asked of the collection team and data that looks out of the ordinary can indicate the need to improve questioning techniques. Both the data collection forms and training methods should be adjusted based upon the trial run's outcome.