Which option would NOT be effective for visualizing trends in a GIS dataset?

Generating a random sampling of features does not provide a meaningful way to visualize trends in a GIS dataset. Trends often relate to patterns or changes over time or space, which require a systematic approach to displaying the entire dataset or specific portions of it to understand relationships and variations.

In contrast, creating a pie chart of attribute data effectively showcases proportions of different categories, allowing for easy comparisons. Using heatmaps assists in visualizing density and identifying hotspots across geographic areas, highlighting trends related to the concentration of features. Applying 3D extrusions effectively represents elevation changes, giving context to terrain and other features that can reflect trends such as population density correlated with altitude.

Overall, while random sampling can be useful for certain analytical purposes, it lacks the systematic approach needed to visualize trends effectively.