User query and lines of code
The article discusses the importance of the relationship between user query accuracy and lines of code. A method of counting lines of code for database queries is proposed and applied to two query languages. ... The relationship between lines of code and user query accuracy is examined with regression models. The results show that lines of code can explain the high percentage of the variance accuracy, with R2 > 0. ... 9 for the entity relationship model query language KQL. ... Chan, “ During empirical studies on database query languages, it is typical to recruit and train subjects who will then be asked to write queries. User performance is usually measured by query accuracy. This approach can be used to study a single query language, as a way to identify the language’s main difficulties for the users, and thus identify appropriate applications to the language. ... Chan further acknowledges that presently there is no clearly defined method of objectively assessing a query. Some languages have implemented various software metrics, which are somewhat effective in measuring user-programming accuracy (851). Why use lines of code as a measure? ... What is the definition of lines of code? ... (1986): “…a line of code is any line of program text that is not a comment or a blank line regardless of the number of statements or fragment of statements on the lines” (852). ... Free format coding allows compounding two or more statements on one line or distributing a single statement to two or more lines. ... project managers count physical lines, 15% count logical lines, and 50% do not count by either method. ... For measuring effort, counting source lines including comments and data is reasonable. For measuring function size, comments and blank lines could be excluded from the calculation (853). ... When lines of code are counted as a measure of productivity. ... To achieve some level of consistency a set of code counting rules have to be prepared in advance and even then, productivity measures based on KLOC/month can penalize high level languages in favor of low-level languages, which require longer codes (Chan 853). ... When the code grows beyond a sub-routine, this LOC measure cannot adequately access program complexity. This observation suggest that lines of code can be used for database queries, this is due to the fact that queries are typically of very small sizes (Chan, 853). The rules for counting lines of code are adapted from the concepts developed by Software Productivity Research (Jones, 1986) for programming languages. With these guidelines all the lines of code are counted i. ... a query with double nesting). ... Grades were based on an assessment of the overall accuracy of the query, considering both syntactic and semantic errors, and their seriousness. ... One may be tempted to generalize, and to state that more lines lead to less accuracy. ... The main reason is that KQL queries in the experiment do not have so many lines of code. ... According to the author this provides very strong evidence that lines of code constitute a very significant determinant for user query accuracy. In conclusion the study has found that lines of code can be successfully used to study query languages. The results show that lines of code constitute an important determinant of user query accuracy. ... It does however point to future possible research on cross-language comparisons with lines of code. Chan states, “ A common criticism of the use of lines of code measures for a programming language is that they are too crude. ... The article discusses the importance of the relationship between user query accuracy and lines of code. A method of counting lines of code for database queries is proposed and applied to two query languages. ... The relationship between lines of code and user query accuracy is examined with regression models. The results show that lines of code can explain the high percentage of the variance accuracy, with R2 > 0. ... 9 for the entity relationship model query language KQL. ... Chan, “ During empirical studies on database query languages, it is typical to recruit and train subjects who will then be asked to write queries. User performance is usually measured by query accuracy. This approach can be used to study a single query language, as a way to identify the language’s main difficulties for the users, and thus identify appropriate applications to the language. ... Chan further acknowledges that presently there is no clearly defined method of objectively assessing a query. Some languages have implemented various software metrics, which are somewhat effective in measuring user-programming accuracy (851). Why use lines of code as a measure? ... What is the definition of lines of code? ... (1986): “…a line of code is any line of program text that is not a comment or a blank line regardless of the number of statements or fragment of statements on the lines” (852). ... Free format coding allows compounding two or more statements on one line or distributing a single statement to two or more lines. ... project managers count physical lines, 15% count logical lines, and 50% do not count by either method. ... For measuring effort, counting source lines including comments and data is reasonable.