Statistically significant differences Sample Clauses
The 'Statistically significant differences' clause defines the requirement that observed differences or results must meet a predetermined threshold of statistical significance to be considered valid or actionable under the agreement. In practice, this means that any claims, decisions, or obligations based on data—such as performance metrics, clinical trial outcomes, or product comparisons—must be supported by statistical analysis demonstrating that the results are unlikely to have occurred by chance. This clause ensures that parties rely on robust, scientifically credible evidence, thereby reducing disputes over inconclusive or random variations and promoting objective decision-making.
Statistically significant differences. Large sample sizes may inflate measures of statistical significance and may lead to false conclusions about the strength of association. The chi-square measure of association, in particular, is susceptible to this possibility. Therefore, the standards for designating whether a relationship can be termed statistically significant have been increased: the ▇▇▇▇▇▇▇’▇ ▇▇▇-square must have probability of a type 1 error of less than .001 and either the Phi coefficient or ▇▇▇▇▇▇’▇ V must have a value of .150 or greater. Throughout this document, any differences reported meet these criteria, unless otherwise stated. ▇▇▇▇▇▇▇’▇ ▇▇▇-square <.001 Phi coefficient or ▇▇▇▇▇▇’▇ V .150 or higher Non-responses have not been included in the analysis. Therefore, throughout this report, unless explicitly stated as a subpopulation, overall results exclude those who did not respond to a particular question.
Statistically significant differences. Large sample sizes may inflate measures of statistical significance and may lead to false conclusions about the strength of association. The chi-square measure of association, in particular, is susceptible to this possibility. Therefore, the standards for designating whether a relationship can be termed statistically significant have been increased: the ▇▇▇▇▇▇▇’▇ ▇▇▇-square must have probability of a type 1 error of less than .001 and either the Phi coefficient or ▇▇▇▇▇▇’▇ V must have a value of .150 or greater. Throughout this document, any differences reported meet these criteria, unless otherwise stated.
Statistically significant differences. Large sample sizes may inflate measures of statistical significance and may lead to false conclusions about the strength of association. The chi-square measure of association, in particular, is susceptible to this possibility. Therefore, we increased the standards for designating whether a relationship can be termed ―statistically significant.‖ The benchmarks shown in Table 5 must be met for us to term an association statistically significant; the ▇▇▇▇▇▇▇’▇ ▇▇▇- square must have probability of a type 1 error of less than .001 and either the Phi coefficient or ▇▇▇▇▇▇’▇ V must have a value of .150 or greater. Throughout this document, any differences reported meet these criteria, unless otherwise stated. ▇▇▇▇▇▇▇’▇ ▇▇▇-square <.001 Phi coefficient or ▇▇▇▇▇▇’▇ V .150 or higher Unlike previous years, non-responses have not been included in the analysis. Therefore, throughout this report, unless explicitly stated as a subpopulation, overall results do not include those who did not respond to a particular question. However, for questions where ―don’t know‖ is a valid response, overall results include those who selected ―don’t know‖ to a particular question, although they are not always shown in a table. Therefore, responses to some questions may not sum to 100%.