5–8 years · Senior — SQL · Excel · Power BI / Tableau · Statistics · Python · Stakeholder reporting
Rejects the null hypothesis; checks the effect size to understand practical significance; considers the context and implications of the finding; evaluates assumptions of the test and checks for Type I and Type II errors; plans next steps for further validation or experiments.
What do you do if the effect size is negligible despite a statistically significant p-value?
Reviews user feedback and identifies pain points; assesses the choice of charts and whether they accurately represent the data; ensures consistent use of colours and labels; checks for unnecessary complexity or clutter; considers the flow of information and logical grouping of visuals.
How do you determine which metrics are most important to display for stakeholder audiences?
Identifies the extent of the issue and the data sources affected; checks the data ingestion process for anomalies; communicates with stakeholders about the potential impact; initiates a rollback to previous data state if necessary; documents findings for future reference.
The issue stemmed from a recent ETL update. How do you ensure this does not occur again?
Assesses the feasibility of the request within the time constraints; prioritises the most impactful metrics to focus on; communicates risks and limitations to stakeholders; begins data extraction and manipulation in parallel to gathering feedback; prepares a backup plan in case of time constraints.
The stakeholders are insistent on having the new metrics included. How do you negotiate the trade-offs?
Identifies the nature of the missing data (missing completely at random, missing at random, or missing not at random); considers using imputation methods such as mean, median, mode, or predictive model; checks for the impact on analysis and whether to exclude affected rows; assesses if it's appropriate to fill with a placeholder or create a new category.
If you decide to impute values, how do you choose the method of imputation?
Details the context of the incident, including how it was discovered; outlines the immediate actions taken, such as data validation or stakeholder communication; explains how the resolution was achieved, including any tools or methods used; reflects on the long-term changes made to prevent recurrence; avoids blaming others for the situation.
What indicators would you monitor earlier to catch similar incidents?
Load six more — you will get questions you have not seen first.
Give me six more →Say your answer out loud before revealing. Reading the answer first feels like learning and is not.