An ML specialist collects daily usage logs and appends customer metadata including age, then creates two categorical features: dayofweek and binned_age. To explore how usage varies by day of week across age groups, which analysis method should the specialist use?

Crosstabulations (contingency tables) are the most appropriate method because they display the frequency distribution of two or more categorical variables simultaneously. This allows the specialist to directly observe how the usage (implicitly represented by the counts within the table) varies across different days of the week and age bins, fulfilling the goal of exploring usage patterns across these groups. A scatterplot is unsuitable as it's typically used for visualizing the relationship between two continuous variables. Word clouds are for text data and would not provide quantitative insights into usage patterns. Boxplots are used to display the distribution of a continuous variable across different categories, but here we are primarily interested in the joint frequency of two categorical variables, not the distribution of a continuous "usage" metric.