Abstract

Many steps are involved in getting data from an experimental unit of an agricultural trial into a final report. Each step may introduce a great variety of errors. Building quality into systems is much more productive than building checks onto the end. Poor quality database have effects on final study results in terms of estimation, significance testing and power; but auditing agricultural trial is a complex process designed to ensure that it will provide a reliable answer to the question being posed. By introducing digit errors into database in a tomato assay, with small sample size, we demonstrate that simple ranges checks allows to detect and therefore correct, the main errors that impact the final study results and conclusions. For investigating significance level and power, two groups of data were simulated, having identical distributions and variances, but different population means. T -tests were carried out and relative frequencies of rejecting Null Hypotheses were determined. We have demonstrate that simple random errors in data affect the conclusions and that some form of data checking is required. Two different methods are analyzed and recommended, exploratory data analysis with and without a second data entry. On the other hand, not all errors that are found by exploratory data analysis are detectable by double data entry.

Keywords

Quality, error rate, double data entry, exploratory data analysis

Creative Commons License


This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS
 
Apr 29th, 4:45 PM

DEVELOPING SYSTEM FOR AUDITING THE DATABASE OF AGRICULTURAL TRIAL

Many steps are involved in getting data from an experimental unit of an agricultural trial into a final report. Each step may introduce a great variety of errors. Building quality into systems is much more productive than building checks onto the end. Poor quality database have effects on final study results in terms of estimation, significance testing and power; but auditing agricultural trial is a complex process designed to ensure that it will provide a reliable answer to the question being posed. By introducing digit errors into database in a tomato assay, with small sample size, we demonstrate that simple ranges checks allows to detect and therefore correct, the main errors that impact the final study results and conclusions. For investigating significance level and power, two groups of data were simulated, having identical distributions and variances, but different population means. T -tests were carried out and relative frequencies of rejecting Null Hypotheses were determined. We have demonstrate that simple random errors in data affect the conclusions and that some form of data checking is required. Two different methods are analyzed and recommended, exploratory data analysis with and without a second data entry. On the other hand, not all errors that are found by exploratory data analysis are detectable by double data entry.