Avni Shah

Who they are

This page records the public professional record only. Nothing adverse is recorded and no motive is attributed.

The affiliation is the one on the 2007 report, Dartmouth College, shared with her co-author. Her wider career was not surveyed for this page, so none is asserted.

Key papers

Shah, A. M., & Wolford, G. (2007), 'Buying Behavior as a Function of Parametric Variation of Number of Choices', Psychological Science, 18(5), 369-370, doi:10.1111/j.1467-9280.2007.01906.x.

The report itself was not opened for this page. What follows rests on press and science-blog accounts of it, and on Barry Schwartz's citation of it in the 2016 edition of The Paradox of Choice.

Key ideas

The accounts agree on the design and the shape of the result. Passers-by were offered a pen to buy, and the number of kinds of pen on display was varied between two and twenty. Purchases were highest at a middle number and lower with both fewer and more options. Dartmouth's George Wolford is quoted as saying the peak was at ten. Schwartz writes that maximum pen purchasing occurred with about eight to ten options (p. 109).

The result is a single study of one cheap product in one setting, and the accounts give no sample size or effect size that could be checked here. It has been read as showing that there is an optimum number of choices, which is a more defensible claim than that more choice is always worse. It does not say where the optimum lies for a shop with thousands of lines.

The finding sits beside the grocery field studies on this site as a number from a different kind of evidence. See choice overload.

Why one short report carries so much weight

A two-page report is an unusual place for a number to come from, and the pen figure has travelled further than the report's length would suggest. It appears in Schwartz's 2016 text and in press coverage, usually without the design. That makes the design worth stating: a range of pens from two to twenty kinds, purchases measured, a peak in the middle.

For a shopkeeper the report is best read as evidence of shape. It supports the claim that there can be too few options and too many, and that the best number lies between them. It is silent on where the peak falls for a shop with a different product, a different price or a different customer.

The report sits beside the grocery field studies on choice overload. Those studies measured sales and household visits across months, and the pen study measured a purchase at a stall. They are different kinds of evidence, and a reader should not average them.

What was and was not read

The report's text sits behind a paywall and was not read for this page. Everything above rests on press and science-blog accounts and on Schwartz's citation. The sample size and the size of the effect are not given in those accounts, so neither is given here. A reader who needs either should obtain the paper.

How the pen figure compares with the grocery figures

The grocery studies on this site produced figures of their own. An 11 per cent average rise in sales across 42 categories as first published in 2001, which the 2005 paper reports with a 2004 correction note as a rise of 1.5 per cent. A quarter longer between orders, and 4.8 per cent less per order, for households offered a reduced range. Each of those comes with a stated sample, a stated setting and a stated limit.

The pen figure comes with a shape and a number and little else. That is not a criticism of the study, which is two pages long, but it does determine how the figure should be used. It is a reason to expect an optimum. It is not a target.

A reader who wants to test the idea in a shop can borrow the design and not the number. Vary the count of options across comparable days or comparable stands, record purchases, and look for a rise and a fall. The pen study suggests what the pattern might look like. Only the shop's own data can say where it peaks.

---