Browse technical information on survey analysis below.

Statistical Calculators - Calculator Icon for confidence intervals

Machine Powered Trend Analysis for I Categories and 2 Response variables

The BIS.Net Team The BIS.Net Team

The tabular output for machine powered alternative for trend analysis, when there are I categories and 2 response variables, one for successes and other for failures, consists of a linear model and a sigmoidal model. The linear model is based on the Cochran Armitage model using weighted regression. The sigmoidal model is based on Logistic Regression. There exist several options for performing significance testing, such as....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

Categorial Ordinal Trend Analysis

The BIS.Net Team The BIS.Net Team

This type of test is used when the row and column variables can be assigned a score from which it is possible to determine if there is a correlation between the row variables and column variables. The scores are usually sequenced from 1 to n, although other scores may be more appropriate. For example consider a study to determine the effect of dosage on pain reduction for cancer patients. Pain levels may be defined on a scale....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

I*2 Categorial Trend Analysis

The BIS.Net Team The BIS.Net Team

I*2 Trend Analysis is extensively used in epidemiology, genetics, political and market research surveys. There are several models used to analyze the trends. Arguably the two most popular are the Cochran-Armitage linear trend and the Logistic Regression function. The latter has become the most frequently used model for binary based data analysis (e.g. proportions). It assumes a sigmoid curve such as as shown below....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

Cochran Armitage Trend Test

The BIS.Net Team The BIS.Net Team

The Cochran–Armitage (CA) test is extensively used in both epidemiology and genetics to test for linear trend in two-way tables with a binary outcome. E.g. Success, failure; positive negative. It can also be used extensively outside these areas, such the service industry, political surveys. As an example consider the following survey to see if there is a relationship between the number of conservative party voters and age groups....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

Kendall’s Tau Statistic

The BIS.Net Team The BIS.Net Team

Kendall’s Tau statistic is used to measure correlation between two sets of ordinal data. There are many applications in marketing, customer service, politics, health where this analysis can be applied. For example, we may wish to compare the performance of two political leaders, asking 50 respondents to rate their satisfaction on a scale of 1 to 10, with 1 being completely dissatisfied, and 10 completely satisfied....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

Predictive value for medical screening and diagnosis

The BIS.Net Team The BIS.Net Team

One objective of Medical Research is to find tests to identify the presence or future development of a disease. For example, the PSA test is used to detect the presence of prostate cancer. Two by Two Contingency tables are often used to study the effectiveness of diagnostic and screening tests. Questions that are answered include how many times do positive test results falsely indicate the presence of a disease....

Continue Reading
Statistical Calculators - Calculator Icon for confidence intervals

Wilcoxon, Mann-Whitney Test

The BIS.Net Team The BIS.Net Team

The Wilcoxon, Mann-Whitney test was designed to compare differences in two independent means, where a normal distribution cannot be assumed. It is called a non-parametric method because it does not consider the actual values, but only the ranks. For example, we may wish to compare the satisfaction with two different political leaders and conduct a micro survey with 6 randomly selected voters. Satisfaction is rated....

Continue Reading
WHO WE ARE

Qtech International has specialized in the development of quality information systems and analytic technologies for leading manufacturing corporations for 30 years in 40 countries. Our research into machine-powered algorithms has spanned over 10 years, with foundation research being university supervised prior. Research has involved many sectors, including manufacturing, health, political campaigning, retail, finance and more.

www.qtechinternational.com