Welcome to Audit Data Analytics
Building an audit community devoted to top-end data analytic approaches.
ADA (Audit Data Assistant) is FREE open-source data analysis software aimed at audits. We are excited to announce that ADA now has a full suite of audit sampling tools and is available for the first time for Mac.
ADA Project Viewer
WHY SHOULD YOU USE ADA?
- ADA is developed and designed by practitioners with 10+ years of experience consulting in audit analytics.
- Our software is really intended to be an easy-to-learn tool that integrates well with Excel and users of Excel. Our software augments your capabilities in Excel. You put the data in. Do what you need to do that you can’t do in Excel. Put the results back in Excel and keep on trucking.
- Your knowledge of Excel carries over because all relevant functions in Excel are available in ADA. Need to get the age difference in days between two dates not counting weekends? We have the Excel function for that (networkdaysintl). We also have other useful functions that are not in Excel such as a fuzzy match similarity function (LevenshteinSimilarity).
- Since you don’t have to learn dozens of strange new functions, you only have to learn how to navigate fewer than 20 dialog boxes to be able to do EVERYTHING.
- Like other audit software, it can handle millions of records and perform operations that cannot be done in Excel.
- Audit Sampling
- Fuzzy Joins and Fuzzy Duplicates
- It also more easily performs operations that can be done in Excel: Appending (copy/paste), Filtering, Summarizing (pivot tables), Sorting, Exact Joins (VLookup) and Exact Duplicates (conditional formatting).
- The development future of ADA is extremely bright! Developed entirely in Python, ADA is on the precipice of being extended into the advanced world of data science and AI techniques.
ADA IS BETTER THAN OTHER SOFTWARE
- When importing from Excel, you can control the resulting data types of columns instead of having the type forced on you. For example, you can force a date column to always come in as a date rather than a date sometimes and a text column at other times. You can save this configuration as a template you can reuse to import multiple Excel files at once and have all the column types be the same.
- PDF Report import is the easiest, most intuitive tool of its kind on the market with an easy-to-follow wizard of screens replete with on-screen instructions that walk you through how to capture the information in a non-flat-file report. You don’t even have to type field names if they already exist in the report. There’s a screen for capturing that. The PDF Report import process flows directly into the Text import process.
- The Excel import and Text import processes are almost mirror images, so learning one process basically ensures you’ve learned the other.
- You can Append datasets where the columns of the same name have different data types. This can be a real headache with other audit software.
- Hate writing date formats? ADA can guess the date format and automatically convert a text column to datetime.
- The Join routine is the most versatile of its type on the market. Other audit software programs perform mainly many-to-one joins and limited many-to-many joins.
- ADA does 4 relationship types of joins (many-to-many, many-to-one, one-to-many and one-to-one) using as many fields as you like.
- ADA can do 6 exact joins as well as fuzzy joins using any of the above 4 relationship types.
- Many-to-one joins will by design ignore viable matches if more than one exists in the secondary database. ADA will warn you in these cases because many-to-one may not be the best technique for the data (you can also turn off the warning).
- If the columns you are joining by (i.e. key columns) are different data types, ADA can oftentimes correctly guess the required conversions and accurately join the data anyway.
- There are 10 procedures for detecting duplicates:
- Exact Duplicates
- Fuzzy Duplicates
- Duplicate Differences where one or more columns have different sets of values while one or more columns have the same value
- Deduplication and reporting of only unique values
- Developed and designed by an M.S. statistician and audit sampling expert practitioner with 10+ years of experience advising hundreds of clients.
- State-of-the-art applications with transparent reporting of all calculations performed.
- Perform Attribute Sampling, Monetary Unit Sampling, Classical Variable Sampling and Random Sampling.
- Plan and Extract samples using only 1 dialog box (not many).
- Designed and built for users to export samples to Excel, audit them for correctness and bring them back seamlessly for quick and easy evaluation.
- All samples can be replicated by using the random seed number.
- All evaluation results including detailed calculations can be quickly and easily exported to Excel for easy report preparation.
- Expands traditional attribute sampling planning to include Error Assurance planning in case a user wants to gauge the likelihood of getting a minimum number of errors, which is very useful in performance audits and variable sampling plans.
Monetary Unit Sampling
- Tired of ending up with samples that are a different size than you planned (usually because of high value items)? With ADA, planning always generates the accurate sample size for the sample you are going to get.
- Calculated sample sizes are accurate and by-the-book, meaning they mirror attribute sampling sizes and do not incorporate any additional unpublished and unjustifiable methods to incorrectly reduce the sample size.
- Calculates an accurate figure for the total allowable taintings one can have in a sample when performing sampling planning.
- Implements the more conservative and industry-standard Stringer Bound evaluation method for low error rate samples by default instead of the Cell Evaluation technique, which according to its inventors Leslie, Teitlebaum and Anderson is intended to be used only as a secondary procedure to the Stringer Bound evaluation.
- Evaluation of low error rate samples provides detailed calculations for each error used in the projection.
- Performs high error rate PPS Sampling Evaluation.
Classical Variable Sampling
- Has the ability to separate both low values and high values from the sampling population so you can audit them 100 percent. Great when auditing General Ledger with large positive and large negative values in the same file!
- Performs the most accurate stratification possible by dynamically defining cells (i.e. potential places to draw a stratum line) for each unique value in the provided population file.
- Evaluation is accurate and by-the-book by following Statistical Auditing by Donald M. Roberts exactly, yielding results with greater precision.
- Evaluation provides detailed calculations by stratum so you can see how all numbers were derived. Want to see the regression estimator’s regression coefficient by stratum or the ratio estimator’s calculated ratio by stratum? ADA will give that to you.