新永資訊 Analyse-it Standard/ Method Evalution 方法評估分析軟體
Analyse-it Method Evaluation edition是一個方法評估分析軟體，檢驗科或者那些需要檢驗、核實分析診斷數據方法的研究員使用
Building on the industry-standard spreadsheet, Analyse-it integrates into Microsoft Excel to provide you with a wide-range of statistics and charts for visualising, describing and testing hypotheses.
Work within the familiar Excel environment
Why learn how to use a complex statistics package? Analyse-it integrates into Excel ’97 to 2007 and provides a toolbar of dataset management and statistical tests so you can analyse data directly from your worksheet. Staying in the familiar Excel environment means there is virtually no learning curve!. You don't have to export your data to another package and you won't end up with your data locked into an in-accessible, proprietary, file format.
Explore, visualise and test sample distribution
Analyse-it provides extensive descriptive statistics and insightful charts for exploring samples. Statistics include the mean, median & standard deviation, all with confidence intervals, plus many non-parametric measures such as quartiles & percentiles.
box-whisker plots, mean-plots, percentile and standard deviation plots let you see how observations are distributed and spot outliers. A frequency histogram, with optional normal curve overlay, provides an alternative view of the sample distribution . Histogram bins can be determined automatically by Analyse-it or you can enter your preferred bins.
Tests and charts include:
- Summary statisticsFrequency histogram with configurable bins
- Box-whisker plots, with near- & far- outliers
- Dot-, Mean-, Percentile- & SD plots
- 1-sample z-test
- 1-sample t-test
- Binomial test
- Runs test
Verify pre-test assumptions of normality
Tests and charts are included for verifying a sample is normally distributed -- a necessary step when using parametric statistics. Hypothesis tests are included so you can formally test normality, or you can use the normal quantile plot and frequency histogram (with normal overlay) to make a visual assessment.
Tests and charts include:
- Kolmogorov-Smirnov
- Shapiro-Wilk (supporting upto 5,000 observations)
- Anderson-Darling
- Normal Quantile plot
- Frequency histogram with normal overlay
Find changes & differences between groups and samples
Parametric procedures including the t-test and ANOVA, with post-hoc comparisons, are provided so you can test whether groups/samples change due to a factor or over time. For non-normal and categorical data non-parametric procedures are provided, including Mann-Whitney, Kruskal-Wallis, Fisher exact, and Chi-Square.
Side-by-side dot-plots, box-whisker plots, & mean-plots, let you visually compare samples. Interaction plots, with connected means, let you easily see whether samples or groups differ or change.
Tests and charts include:
- Independent & paired t-test
- Mann-Whitney & Wilcoxon Signed ranks
- 1-way ANOVA, with LSD, Bonferroni, Tukey, Dunnet & Scheffe post-hoc comparisons
- 1-way repeat-measures & 2-way ANOVA
- Kruskal-Wallis & Friedman ANOVA
- Sign & Median test
- F- test
- Fisher exact & Chi-Square
- McNemar & Cochran tests
- Side-by-side dot-plots, box-whisker plots, mean-plots & SD plots
- Interaction plot (connected means)
Identify trends and relationships to make predictions
For bivariate data, Analyse-it lets you explore potential trends with Pearson, Spearman, and Kendall correlation, and compare categorical ratings with Chi-square & Kappa.
Where definite trends exist, the variables predicting the response can be identified with multiple linear regression and polynomial regression. Goodness of fit statistics and residual plots tell you how well the chosen variables predict the response. The equation to predict future observations is also shown.
- Single & multiple linear regression
- Polynomial regression
- Scatter plot with fit, fit confidence and prediction bands
- Raw and standardised residual plot
- Histogram of residuals
- Pearson, Spearman correlation
- Kendall concordance
- Kappa & Weighted Kappa
- Bland-Altman agreement
Publication-quality print-ready reports
Analyse-it's perfectly formatted reports mean you can print the analysis immediately, without wasting time tidying them up for presentation. You can also copy-paste charts and statistics directly into your own documents and presentations. Chart look is based on the most popular formats used in published research, so charts are immediately ready for inclusion into research papers.
Save, exchange and update your analyses
Once analysed, your Analyse-it analyses can be saved in a standard Excel workbook, published to your intranet, or e-mailed to colleagues. Colleagues with Excel can open and print workbooks containing Analyse-it analyses, without needing Analyse-it. At any time you, or colleagues with Analyse-it, can update the analysis reports, repeating the analysis to reflect changes or additions to the source data.
A proven reliable product
We've sold and developed Analyse-it for over 11 years now, and it's used by over 15,000 customers world-wide (see selected customer list). It's widely respected by researchers and cited in well over 1,500 published peer-reviewed papers in internationally recognised journals. As a company we've developed statistical software for over 16 years, developed custom statistical software for two of the world's largest pharmaceutical companies and have worked in the software industry for over 25years.
Some statistical software providers might try convince you Excel is unsuitable for rigorous statistical analysis. With Analyse-it those fears are unfounded. Analyse-it doesn't use any of the built-in Excel functions. Instead all Analyse-it's statistical tests are written in C++ for lightening fast calculation, then rigorously tested to ensure they are accurate.
Full support and help along the way
If you need help or advice on using Analyse-it, the full help and tutorials included with Analyse-it provide easy-to-understand, plain English instructions to guide you through using Analyse-it. Help on choosing a suitable statistical test, then using, interpreting, and understanding the statistics is included. Of course should you get stuck, our widely-praised customer support are ready to help, with most queries are answered within a few, and almost all within 24, hours.