In ranking questions, lower scores are better. The item with the lowest rank is the one respondents treated as most important.
The Four Columns
Average Rank
The mean rank score across all respondents for that item. The item with the lowest Average Rank was ranked most important most consistently across the sample.
This is the primary column for reporting. When comparing items or groups, start here.
Median Rank
The middle rank value when all individual responses for that item are sorted in order. Less sensitive to outliers than the average.
Use Median as a validation check alongside Average Rank. If the two figures diverge noticeably, the data may be skewed by outliers and is worth investigating before reporting.
Top Rank
The single best, or lowest, rank assigned to that item by any one respondent. If any respondent ranked the item first, Top Rank = 1.
This is not an average. It is not a majority view. It reflects one person's response at the ceiling of enthusiasm for that item.
Lowest Rank
The single worst, or highest, rank assigned to that item by any one respondent. It shows how unfavourably the item was viewed by its least convinced respondents.
Like Top Rank, this is a single-respondent extreme, not a summary of the group.
Reading All Four Together
Each column tells you something different:
- Average Rank: where most respondents landed
- Median Rank: a sanity check on the average
- Top + Lowest Rank together: the spread of opinion
A narrow gap between Top and Lowest signals consensus. Respondents broadly agreed on how important the item is. A wide gap signals a polarising item, where some respondents ranked it highly and others ranked it poorly.
Narrow spread means consensus. Wide spread means polarization. Average Rank tells you where the group landed overall, but the spread tells you how much agreement sits underneath that number.
Age-Split View
When results are broken out by age group, the same four columns appear for each group. Apply the same reading logic: Average Rank is the primary comparison column within each group, and Top-Lowest spread still indicates how polarising the item is within that group.
Worked Example
Consider a survey where respondents ranked five product features in order of importance.
| Feature | Average Rank | Median Rank | Top Rank | Lowest Rank |
|---|---|---|---|---|
| Fast delivery | 1.4 | 1 | 1 | 3 |
| Loyalty rewards | 2.9 | 3 | 1 | 5 |
| Eco packaging | 4.1 | 4 | 2 | 5 |
Fast delivery: lowest Average Rank, narrow Top-Lowest spread from 1 to 3. Strong consensus that this feature is the most important. Lead with this in reporting.
Loyalty rewards: moderate Average Rank, full spread from Top to Lowest from 1 to 5. Some respondents ranked it first, while others ranked it last. This is a polarising item, so average alone does not capture the picture. Note the spread when reporting.
Eco packaging: high Average Rank, also a full spread. It was consistently ranked as less important on average, but the spread shows at least one respondent ranked it second. The average here is reliable because Median Rank (4) and Average Rank (4.1) are close, confirming the average is not being pulled by outliers, but the item is not a reporting priority.