# [Video] DAT Data Validation

DAT iQ Benchmark Training Tutorials - 6 - YouTube

Data validation is crucial for both new and experienced users to comprehend fully the information they are working with within the application.

## Getting Started with Data Validation

As an analyst, understanding the data being reported is paramount. Our application offers several features to aid in this understanding:

- **Trends Analysis:** Utilize the trend object to analyze data trends, such as loads per week or year. This helps confirm that the data matches expectations.

- **Equipment Types:** It's important to note the equipment types currently reported (e.g., Dry Van, Flatbed, Intermodal, Reefer) and recognize excluded types (LTL, Ocean, Parcel, Rail, etc.).
- **Cost Analysis Over Time:** Analyzing cost over time offers valuable insights for financial planning and strategy.

- **Recent Data Examination:** Leveraging the lanes object to sort and examine recent data can highlight significant trends or outliers that merit further investigation.

## Deep Dive into Data

Validating your company’s data involves several practical steps:

- **Verify Expected Records:** Ensure the records in the data match your expectations, including metrics like loads per year.
- **Check for Outliers:** Identify any significant deviations, such as exceptionally high or low rates, and verify their accuracy.
- **Understand Inclusion and Exclusion:** Confirm the types of data included and excluded in reports, such as equipment types and transportation modes.
- **Carrier Names Scrutiny:** Scrutinize carrier names to catch any unexpected entries, ensuring no irrelevant data slips through.

## Leveraging Built-In Validation Tools

DAT iQ Benchmark includes built-in validation features to streamline the data verification process:

- **Default Views:** The default view showcases data deemed clean and reasonable, based on extensive backend testing.
- **Validation Tests:** Automated tests check for logical consistency, such as matching cities with states and appropriate distance or rate figures.
- **Legacy and AI Model Enhancements:** For legacy customers, familiar tests are still in place, complemented by improved tests from newer AI models, offering a layered approach to data validation.

## Addressing Data Outliers

Upon detecting outliers, users can:

- **Update Page Views:** A quick update can reveal outliers in the current dataset.
- **Analyze and Report:** Download detailed spreadsheets for a deeper analysis of outliers, including explanations and comparisons to benchmarks.