PL-300 Visualize and analyze the data Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  CALCULATE(COUNTROWS(Sales), DATESINPERIOD('Date'[Date], MAX('Date'[Date]), -30, DAY))

  Why this is correct

  This expression is correct because DATESINPERIOD constructs a continuous, dynamic 30-day window ending at the boundary defined by MAX('Date'[Date])—the latest date present in the current filter context. By specifying DAY as the interval type and a negative offset of -30, the function returns the set of dates from MAX('Date'[Date]) minus 29 days through MAX('Date'[Date]), inclusive of both endpoints, resulting in exactly 30 calendar days. Since it is anchored to the maximum date in the date table rather than to a hardcoded system date like TODAY(), it remains accurate even when the underlying data has not been refreshed up to the current day, and it automatically adapts to whatever date range is selected in a report slicer or page filter, making it the most robust choice for a rolling last-30-days measure.

• ✗

  CALCULATE(COUNTROWS(Sales), DATESBETWEEN('Date'[Date], TODAY()-30, TODAY()))

  Why it's wrong here

  This expression is incorrect because DATESBETWEEN creates a fixed window from TODAY()-30 to TODAY(), which assumes both dates exist in the 'Date' table and that 'today' is the reference point for the desired period. In practice, the date table may only span through the last data load date, which could be earlier than the system's current date, causing the filter to evaluate over an empty or incomplete date set and produce a blank or understated result. Furthermore, this approach does not respect the filter context of a report—if a user filters to a specific historical month, the measure still ignores that selection and returns sales for the 30 days leading up to the real-world today, making it inconsistent for interactive analysis where the expected behavior is to show the trailing 30 days relative to the selected period's end date.

• ✗

  CALCULATE(COUNTROWS(Sales), PREVIOUSMONTH('Date'[Date]))

  Why it's wrong here

  This expression is wrong because PREVIOUSMONTH returns the entire calendar month that precedes the month of the current filter context, not a rolling 30-day interval. For example, if the current context is May 2025, this function returns all dates from April 1 to April 30, regardless of whether the user is viewing a specific day, week, or year-to-date selection; the number of days can vary between 28 and 31, so it does not reliably represent a 30-day span. Additionally, PREVIOUSMONTH ignores the day-level granularity needed for a trailing window—it resets to month boundaries and cannot reflect the 30 days immediately preceding an arbitrary date such as the 15th of the month, making it inappropriate for any kind of 'last 30 days' metric.

• ✗

  CALCULATE(COUNTROWS(Sales), DATEADD('Date'[Date], -30, DAY))

  Why it's wrong here

  This expression is incorrect because DATEADD with an offset of -30 days shifts the existing set of dates in the current filter context back by exactly 30 days, rather than isolating the trailing 30-day range ending at the maximum date. If the current filter context contains a single date, DATEADD returns only that one date shifted back by 30 days, which would count rows for just one day, not a 30-day window; if the context contains a full month or year, DATEADD returns that entire period shifted backward, producing a comparison of the same-length period shifted in time but not bounded by the latest date. This function is designed for period-over-period comparisons (e.g., comparing this month to the same month last year) and does not create the dynamic, end-anchored window that DATESINPERIOD provides.