Glossary “T” Definition of Terms in DataMyte
t-Distribution
Used for a sample with size n, drawn from a normally distributed population, with mean x and standard deviation s. The true population parameters are unknown. The t-distribution is expressed as a table for a given number of degrees of freedom and α risk. As the degrees of freedom get very large, it approaches a z-distribution.
t-Test
A test of the statistical hypothesis that two population means are equal. The population standard deviations are unknown, but thought to be the same. The hypothesis is rejected if the t value is outside the acceptable range listed in the t-table for a given α risk and degrees of freedom.
Tolerance
The permissible range of variation in a particular dimension of a product. Tolerances are often set by engineering requirements to ensure that components will function together properly.
Total Indicator Readout (runout)
Measurement of a characteristic such as concentricity or flatness for maximum or minimum value as well as range. The part is rotated or moved such that the gauge measures the entire area in question. The value(s) of interest, such as maximum, are either read manually from the gauge or determined automatically by automated data collection equipment.
Total Quality Control (TQC)
A management philosophy of integrated controls to ensure customer quality satisfaction and economical costs of quality. Includes engineering, purchasing, financial administration, marketing, and manufacturing.
Transcription
Rewriting; copying.
Trend
A gradual, systematic change with time or other variable.
True x Causes
For x control charts: changes in the x control chart which are due to actual changes in the mean produced by the process. True x changes are usually accompanied by a stable pattern in the R chart.
TurboNET ™
SPC networking software that joins DataMyte data collectors to a personal computer running TurboSPC software.
TurboSPC ®
Quality management software for the PC. TurboSPC software receives data from data collectors, other programs, or keyboard inputs. The software produces control charts, performs capability analyses, creates reports, and provides a versatile database and analysis tools for quality management functions.
TurboSPC Observe
TurboSPC Observe works with Novell® Netware to link SPC data from a central work station with remote locations.
Type I Error
In control chart analysis: concluding that a process is unstable when in fact it is stable.
Type II Error
In control chart analysis: concluding that a process is stable when in fact it is unstable.