Premium
Data analysis for non‐statisticians
Author(s) -
Neuman R. C.
Publication year - 1994
Publication title -
journal of vinyl technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.295
H-Index - 35
eISSN - 1548-0585
pISSN - 0193-7197
DOI - 10.1002/vnl.730160205
Subject(s) - outlier , computer science , data mining , identification (biology) , process (computing) , sequence (biology) , data cleansing , data analysis , artificial intelligence , data quality , engineering , metric (unit) , operations management , botany , genetics , biology , operating system
Experiments generate data. Data generate results. Results lead to decisions. Decisions solve problems. Right??? Well, not always. Things can go wrong…. At the root of the above sequence is data. If the data are not good, the problem is unlikely to get solved. If the data are good, but not properly interpreted, the problem may persist. One cause of poor data is a poor experimental design. Once run, the results cannot be improved by the best data analysis techniques. Given good data, interpretation can change the conclusions. Data analysis techniques are available that enable the experimenter to Correct for process drift during a series of experiments. Study data for outliers, which can falsely influence results. Determine distribution of data within and between data sets. Do regression analysis to establish mathematical relationships.These techniques lead to statistically correct identification of cause and effect relationships, or show that none are present in the data. Rigorous data analysis produces fewer disappointments and more problem solutions.