Free CompTIA DA0-002 Actual Exam Questions

Good point on permissions not usually filtering by date. I think C could be ruled out because a refresh failure might just show old data, not restrict to current month only. B sounds right here.

D imo, connectivity issues could stop full data loads, limiting results to current month.

I see why AND (C) is popular here, but another angle is that NOT (B) could be useful to invert or test conditions negatively, depending on what “evaluating all conditions” means. Still, if the goal is to check all conditions together as true, AND is the straightforward choice. OR and IF don’t really fit if you want to ensure every condition passes. So yeah, C makes the most sense for verifying all conditions are met.

It’s C because AND ensures every single condition is true, no exceptions.

D imo, because skewness describes asymmetry, not the range of the middle data. That rules out D easily, narrowing it down to A since none of the others relate to the middle 50%.

Not B, because the difference between mode and median doesn’t focus on spread or range; A is better since the interquartile range actually captures the central 50% of data points.

Maybe D makes the most sense since UAT is about making sure users actually get what they need, not just that the software runs without bugs or integrates well, which are covered earlier in testing stages.

D UAT is all about confirming the software meets user needs before launch.

D imo, this feels more like skewing because the data distribution suddenly became unbalanced with that viral spike. The model likely was trained on more stable, evenly spread demand, so now that one type of data dominates, it’s not predicting well. Data drift usually implies a gradual change over time, but here it’s a sudden, extreme shift causing the supply chain issues.

B/C? The demand spike definitely changes the data pattern, pointing to data drift, but you could argue the model wasn’t sized for this extreme case either. Both seem plausible depending on perspective.

Probably C again, but thinking differently: standardization (A) just scales the data and doesn’t actually split customers into groups. Concatenate (B) and appending (D) are more about combining datasets, not segmenting. So if the goal is to create clear customer groups based on spending, binning (C) is the only option that actually cuts the continuous spending data into segments. It’s the only one that directly addresses grouping rather than just transforming or merging data.

Good point about binning, but also think about what concatenation (B) and appending (D) do—they just combine data, so they won’t help with segmenting customers by spending. Standardization (A) normalizes the data but doesn’t create groups on its own. So really, binning (C) is the only technique that turns continuous spending values into distinct segments, which fits the goal perfectly.

A/D are wrong because = NULL or = 'NULL' doesn’t work for actual NULL values in SQL. C is the opposite of what’s needed. B is the standard way to check for real NULLs without mixing in empty strings.

B is best, but just make sure the empty fields aren’t blank strings instead of NULLs.

Makes sense to use JSON since APIs output it natively, so B.

It’s B for sure. APIs usually give you JSON directly, and pandas has built-in support to read JSON into DataFrames easily. Even if the database isn’t originally JSON, you can convert or export it without much hassle. Options like varchar, TXT, or just strings don’t naturally fit API responses or structured data merging. JSON keeps things clean and structured for merging in pandas.

Not B, the CFO probably doesn’t want to sift through lots of text—something concise like D makes more sense for quick decision-making.

D fits best since CFOs usually want a quick summary, not all the raw data.

B imo, SUM needs GROUP BY when selecting other columns.

Guessing B because without GROUP BY, you can’t use SUM with other columns like firstName and lastName. Fixing just SORT BY to ORDER BY won’t solve the aggregation error.

It’s C and E because triggers protect data that’s both sensitive and restricted access.

Maybe C and F make sense too. If the data is sensitive, it often needs encryption, so the triggers might be for encrypted info as well. Encrypted isn’t really a classification like public or private but more a protection method, so it could be relevant here alongside sensitive. That way, you cover both the type of data and its security state, which aligns with using security triggers to catch unauthorized access or changes. So I’d say sensitive data definitely, plus either private or encrypted depending on how you interpret the classifications.

Maybe B. KPIs are designed to measure things like wait times directly, so they’d give clear data on where delays happen without needing to guess from surveys or visuals.

Maybe A, visuals can quickly reveal where delays happen in the support process.

A imo, since birth month and DOB repeat the same info on different levels.

I’m thinking it’s more about redundancy since DOB and birth month are basically the same info repeated in different detail; duplication would need actual repeated records, which we don’t know for sure. A

I’d put the heat map in the biggest spot since county outages cover the whole state—makes it easier to spot problem areas quickly. The bar chart for status fits well in a smaller section, with a straightforward color scheme like red and green to show outage vs restored states.

Line charts are best for trends because they show changes over time clearly.

C This is definitely a scenario where union fits best since you’re just combining two sets of rows with the exact same columns. Join would be pointless here because there’s no relationship to match on between the two tables—they’re just separate groups that need to be listed together. Group and nested don’t really apply either since those are more for aggregating or subquery stuff, not straightforward combining. So union is the simplest and cleanest way to get all managers and non-managers in one combined report.

This one feels like a clear case for C. Union is designed to combine rows from two tables with the same structure, and since both have matching columns, stacking them makes sense. Joins would only work if we needed to link related data, which isn’t the case here since managers and non-managers are just separate groups. So, C fits best.