Real Statistics Release 6.5

I am pleased to announce Release 6.5 of the Real Statistics Resource Pack. The new release is now available for free download at Download Resource Pack for Excel 2010, 2013, 2016, 2019 and 365 in both Windows and Mac environments.

The Basics, Regression 1, Correlation and Reliability and Time Series examples workbooks have also been revised for compatibility with the new release. These are available for free download at Download Examples Workbooks.

Over the course of the next several days, the website will be updated for compatibility with the new release.

If you are getting value from the Real Statistics website or software, I would appreciate your donations to help offset the costs of the website by going to Please Donate.

The following is an overview of the new features in Release 6.5.

New Histogram with Normal Curve Overlay data analysis tool

A new Histogram with Normal Curve Overlay data analysis tool has been added. This tool provides all the capabilities of the previous Frequency Table data analysis tool for input in raw data format, plus it allows you to optionally overlay the histogram that is displayed with a normal curve that approximates the data. Two options are provided: (1) the area under the normal curve is approximately the same as the area of the histogram or (2) the maximum height of the normal curve is the same as that of the histogram. Generally, option (1) should be chosen.

Revised Frequency Table data analysis tool

The existing Frequency Table data analysis tool has been replaced by the Frequency Table Descriptive Statistics data analysis tool. The new tool contains all the capabilities of the old tool for input in the form of a frequency table (i.e. descriptive statistics and ability to convert the frequency table data into raw data format). In addition, the data analysis tool now gives you the option of performing a Kolmogorov-Smirnov test for normality on the data in the frequency table.

Rasch Item Response Enhancements

The Rasch Item Response data analysis tool has been enhanced to support the PROX method in addition to the UCON method introduced in the previous release of the Real Statistics software. The PROX method supports Likert scores. The data analysis tool now also provides the ability to remove any subject or item with either a perfect or zero total score.

In addition the following new functions have been added to support the UCON method. Here, R1 contains an array of scores where rows correspond to subjects and columns to items (w/o row and column headings), R2 is an array with two columns (column 1 contains subject labels and column 2 contains ability parameter values) and R3 is an array with two columns (column 1 contains item labels and column 2 contains difficulty parameter values).

RASCH_EXP(R1, R2, R3): returns an array with the expectation values (as for the RASCH function)

RASCH_FIT(R1, R2, R3): returns an array with the fit values (as for the RASCHFIT function)

For the following functions, R1 is as above except that if head = TRUE (default) then R1 contains row/column headings and if head = FALSE then R1 is w/o headings.

RASCH_SUBJ(R1, head): returns an array with three columns consisting of the subject labels and corresponding ability parameters and standard errors based on the data in R1

RASCH_ITEM(R1, head): returns an array with three columns consisting of the item labels and corresponding difficulty parameters and standard errors based on the data in R1

For the following function, R1 is as above (with headings).

RASCH_INIT(R1): returns output similar to that in R1, but with items and subjects with all or no correct answers eliminated

The following functions apply the PROX method to the data in R1.

PROX_SUBJ(R1): returns a summary of the PROX method for the data in R1 (with headings) consisting of 4 columns: subject labels for each of the subjects in R1, corresponding ability parameter estimates, standard errors and % correct scores.

PROX_ITEM(R1): returns a summary of the PROX method for the data in R1 (with headings) consisting of 4 columns: item labels for each of the items in R1, corresponding difficulty parameter estimates, standard errors and % correct scores.

PROXX_SUBJ(R1, R2): returns a two-column array consisting of ability parameter estimates and standard errors corresponding to the subject scores in the column array R2 based on the PROX method for the data in R1 (w/o headings).

PROXX_ITEM(R1, R2): returns a two-column array consisting of difficulty parameter estimates and standard errors corresponding to the item scores in the column array R2 based on the PROX method for the data in R1 (w/o headings).

New regression data analysis tools to identify and correct for autocorrelation

The following data analysis tools have been added to the Reg tab.

Breusch-Godfrey and Newey-West: conducts the Breusch-Godfrey test for autocorrelation and calculates standard errors of the regression coefficients that corrects for autocorrelation using the Newey-West approach.

Cochrane-Orcutt Regression: modifies the OLS regression coefficients and standard errors to correct for autocorrelation using the Cochrane-Orcutt approach.

In addition, the chi-square version of the Breusch-Godfrey test has been revised. This affects the BGSTAT and BGTEST functions.

Mann-Kendall and Sen’s Slope data analysis tool

A new Mann-Kendall and Sen’s Slope data analysis tool has been added to the Time Series tab which conducts the Mann-Kendall test and calculates Sen’s Slope.

Changes for the Mac version

The name of the Mac 2016/2019/365 version of the Real Statistics Resource Pack has been changed to XRealStats Mac.xlam. This will cause the Real Statistics add-in to load after the Solver add-in. This change is necessary for some Mac configurations based on recent changes made by Microsoft to Excel. A similar change was made to the Windows version in the previous release of the Real Statistics software.

Existing users of the Mac version of Real Statistics will need to de-install the old version of the software and then install the new version. To de-install the old version of the software, select Add-Ins from Excel’s Tools menu and uncheck Realstats Mac 2016. You can then install XRealStats Mac in the usual way, as described at Real Statistics Resource Pack for the Mac.

Note one other thing, When you open a previously created Excel workbook that references Real Statistics functions, you will receive a message similar to

“This workbook contains links to other data sources”

You then need to follow the installation instructions at Real Statistics Examples Workbooks.

None of these changes apply to the Excel 2011 version of the Real Statistics software.

Enhancement to certain table lookup functions

The following table lookup functions have been enhanced as described below.

LPROB, KSPROB, KSTEST, KS2PROB, DPROB, QPROB, ADPROB

The output from these functions is based on a corresponding table of critical values. For p-values that are smaller than the smallest alpha value in the table zero is returned and for p-values that are larger than the largest alpha value in the table a one is returned.

With this release, you can choose to instead return a value such as “< .01” to mean that the p-value is less than .01, the smallest alpha value in the table of critical values (instead of 0), or “> .2” to mean that the p-value is greater than .20, the largest alpha value in the table (instead of 1). The actual numeric values (.01 and .2 in the above examples) depend on the particular table of critical values.

This new capability is implemented by appending a new argument to each of the functions. When this argument is FALSE (default) the function returns a 0 or 1 if necessary, but when TRUE the function returns a text string of the type described above. This new argument is the last argument for each of the above functions.

Bug Fix

The SSAnova2(R1, r) function now works properly when the data in R1 is in Excel format with headings (i.e. when r > 0).