Smith's Salience Index (S)
Also known as: Smith's S, Salience Index S, Free-List Salience, Item Salience Score
Smith's salience index, conventionally written S, is the standard statistic for summarizing free-list data: for each item it combines how many informants mentioned the item with how early the item appeared in their lists. Within a single list each item receives a local salience equal to the number of items below it divided by the list length, so the first item scores highest and the last scores lowest; S is then the average of that local salience across the entire sample, counting zero for informants who never listed the item. The result is a single per-item number that ranks the items of a cultural domain by their joint frequency-and-priority prominence.
Key highlights
- Reduces an entire free-list dataset to one interpretable per-item number that orders the domain by cultural importance.
- Fuses mention frequency and recall priority, so it outperforms raw frequency at identifying prototypical items.
- Normalized by each informant's list length, making scores comparable across people who list few or many items.
- Directly implemented in standard tools (ANTHROPAC, AnthroTools), making the workflow reproducible and easy to report.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use Smith's S whenever you have rank-ordered free-list data and want a single defensible number that orders the items of a domain by cultural importance rather than by bare frequency. It is the natural summary statistic before selecting items for pile sorts, triad tests, or consensus surveys, and for comparing the salience profiles of subgroups. It requires that mention order was actually recorded and that list lengths are not pathologically short or uneven. It is the wrong tool when only presence/absence was collected (use frequency), when you need relations among items rather than their prominence, or when lists were elicited under heavy probing that destroyed the natural recall order.
Strengths & limitations
- Reduces an entire free-list dataset to one interpretable per-item number that orders the domain by cultural importance.
- Fuses mention frequency and recall priority, so it outperforms raw frequency at identifying prototypical items.
- Normalized by each informant's list length, making scores comparable across people who list few or many items.
- Directly implemented in standard tools (ANTHROPAC, AnthroTools), making the workflow reproducible and easy to report.
- Conflates two distinct signals — being common and being recalled early — so a high S can arise from either.
- Assumes recall order is meaningful, which fails if the elicitation used heavy probing or category prompts.
- The linear position-to-salience mapping is a convention; alternative weightings can reorder mid-ranked items.
- Inherits all coding decisions: over- or under-merging items during normalization changes the scores directly.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How is Smith's S different from just counting how often an item is mentioned?
Frequency counts treat every mention equally regardless of where it appeared in a list. Smith's S weights each mention by how early it occurred (normalized by list length) and then averages over all informants, including zeros for non-mentioners. Two items mentioned equally often can therefore receive very different S values if one is consistently recalled first and the other last.
What value can S take and how do I read it?
S ranges from 0 to 1. An item listed first by every single informant would approach 1; an item nobody lists scores 0. In practice the highest-salience items in a domain sit well below 1, and you interpret S comparatively — the ordering of items and the gap between the high-salience head and the long low-salience tail matter more than any absolute threshold.
Do I need to record the order in which informants listed items?
Yes. The position term (L_i − r_ti + 1)/L_i is the heart of the statistic, so mention order must be preserved during elicitation. If you only captured which items were mentioned, you can compute frequency but not a meaningful Smith's S, and you should report a frequency-based salience instead.
Sources
- 1.Bernard, H. R. (2017). Research Methods in Anthropology: Qualitative and Quantitative Approaches (6th ed.). Lanham, MD: Rowman & Littlefield.ISBN 9780759112421
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Smith's Salience Index (S). ScholarGate. https://scholargate.app/anthropology/freelist-salience-smiths-s