Msc Study By Dawn Anderson December 2014
Whilst popular historical thinking has led many to believe that there can never be too much choice for consumers, recent studies show that there may be negative effects of both too much and not enough choice. Is this ‘too-much-choice’ effect reflected in both offline and online consumer environments, how much is just enough choice, and do consumers display aversion to single option items?
Introduction and Background
Experiments with rats, pigeons, monkeys and humans indicate that organisms prefer choice over no choice (Catania, 1975), (Catania and Sagvolden, 1980), (Suzuki, 1999), (Voss and Homzie (1970), even if the choice outcome provides benefits only equal to those without choice. In their paper entitled “The Lure of Choice”, Bown et al conclude, “the inherent attractiveness of choice, even when it is disconnected from any ultimate benefits, leads retailers to offer it and consumers to be lured by it.” (Bown et al. 2003).
Popular historical thinking speaks of human needs for ‘self-determination’, ‘automony’, and ‘freedom of choice’ and the benefits derived from these across motivation and task-performance aspects (Zuckerman et al. 1978). Furthermore, past thinking has largely believed there can never be too much consumer choice; that those choosing have an infinite capacity to deal with ever increasing options.
But is this true on either point? Is more choice better and human capacity to choose unlimited? Some researchers have referred to a limit to the human capacity to process information and less than optimal ‘heuristics’ people develop when faced with decision making (Miller, 1956), (Simon, 1955), (Schwartz,2004), (Wright, 1975), (Tversky, 1972).
Whilst the initial attraction of variety (Iyengar and Lepper, 2000), (Bown et al. 2003) may prevail; some researchers claim ‘over-choice’ (too much choice) can lead to demotivation and lack of action at the point of purchase (the choice) in consumer buying behaviour, a lack of motivation in task completion, reduced task performance (Iyengar and Lepper, 2000), dissatisfaction and post-choice regret (Iyengar and Lepper, 2000), (Iyengar et al. 2006). Findings also indicate that subjects simply choose the easiest option to understand in larger, when compared with smaller, choice sets (Iyengar and Kamenica 2010).
Others argue it is not the number of items that make too much choice an issue but the number of attributes across those items to consider – the complexity of those choices (‘choice complexity’), regardless of the size of the choice set (items) – (Greifeneder et al. 2010).
But when does too much choice occur, and is this ‘too-much-choice’ effect reflected online as well as in offline consumer environments? If, as Schwartz in his ‘Tyranny of Choice’ paper states, echoing Tversky, (Tversky, 1972) “As the number of choices we face increases, the psychological benefits we derive start to level off… Some of the negative effects of choice… begin to appear and rather than level off, they accelerate. (Schwartz, 2004)”, where is that point?
Whilst studies carried out have compared limited ranges (2-6) versus extensive ranges (24-30), and others have looked at small number comparisons (typically between 2-6), few have looked at the range between the two extremes.
If ‘choice-complexity’ (Greifeneder et al. 2010) also plays a part, can simplifying the choice process for limited-capacity consumers help with ‘over-choice’ in larger choice set presentation offline and via website functionality online, given past research suggests people may defer choice when presented with a difficult task of choosing (Dhar, 1997), (Iyengar and Lepper, 2000)?
As Chernev states “Those without an idea of what they wish to choose beforehand, face a very complex task indeed of not only searching through an assortment but also of compiling their ‘ideal point availability’ (combination of attributes of an item they wish to choose)” (Chernev 2003).
At the other extreme, can showing fewer rather than more options also have negative effects leading to choice deferral? Recent research claims displaying single options can lead to ‘single- option-aversion’ (Mochon 2013) – whereby consumers admittedly prefer an option but still have an innate desire to search, displaying adversity for a single option presented alone. Once exposed to the single-option-aversion’ phenomenon, the desire to search carries forward to future considerations, delaying choice further, claims Mochon (Mochon 2013).
Identified Issues And Their Impact
The issue of whether the ‘too much choice’ effect exists, and, if so, at what point, is an important one for retailers when considering how best to present products for consideration either offline or online.
If, presenting too many options via larger choice sets produces only a browsing effect or the ‘Lure of Choice’ (Bown et al. 2003), (Iyengar and Lepper, 2000), over a buying-effect (demotivation to commit to choosing) (Iyengar and Lepper, 2000) then there are potentially serious implications for industry. It could be argued that retailers could sell more by displaying less, or even reducing product range sizes and variations.
Conversely, if consumer adversity toward a single-option display exists and continues following initial exposure to lone item presentation, then retailers may well be mindful to decide whether showing items simultaneously or sequentially would be most beneficial. Likewise, consideration would be justified as to whether alternatives should always be displayed with items.
If the ‘too much choice’ phenomenon occurs as a result of complexity rather than mere simultaneous item volume then policy makers may need to consider how best to assist the consumer in choosing, via categories, product filters and search facilities online and their best available offline equivalent.
When is there ‘just enough choice to consider’ simultaneously (‘an optimum number of display items’)? If there exists a point at which the positive benefits available from choice begin to plateau and move towards a negative point, where is this? Is this ‘just enough choice’ number reflected both online (on a webpage) and offline (store or supermarket) when people have to choose between options? When viewing lone items (they confirm would be acceptable), do people actively choose to seek alternative options or do they proceed to buy without further search?
In a 2000 study by Iyengar and Lepper, ‘too much choice’ was investigated comparing 6 items (limited range) with 24 items (extensive range). No mid-point ranges were considered (12, 18 items) (Iyengar and Lepper, 2000). This study aimed to look at whether additional sized ranges of 12 or 18 between these two limited (6 items) and extensive (24 items) numbers produced different results where an ‘optimum’ ‘tipping’ point becomes ‘too much choice’.
The study also aims to consider whether the numbers of products participants preferred to view simultaneously offline in a retail environment was reflected in an imaginary online ecommerce environment.
The study also sought to identify whether participants felt a need to seek alternatives to compare to a product which was acceptable (they liked the product), simply for the sake of having a comparison and because it was a lone item. This was in response to recent research hypothesising that consumers hold an adversity for a single item and will defer choice if a single option is available – ‘single option aversion’ (Mochon 2013).
To identify how many participants would indicate they preferred to compare 6, 12, 18 or 24 items in an assortment simultaneously in a shop or supermarket (NOT A WEBSITE). (limited and extensive range (6 and 24)) and two middle sized assortment ranges (12, 18) and whether there was a clear preference.
To identify how many items participants indicated they would feel comfortable comparing simultaneously on a website page and whether they preferred comparing larger numbers on a website page and / or found it easier than comparing larger numbers in a shop or supermarket.
To identify whether participants confirmed they would choose not to buy a satisfactory lone product they liked because they preferred to always compare to alternatives – i.e. would they display ‘single- option aversion’ (Mochon 2013)?
Past Research and Findings (Literature Review)
Empirical research concludes that humans, as organisms, along with other species such as rats (Voss and Homzie, 1970), monkeys (Suzuki, 1999) and pigeons (Catania, 1975), Catania, and Sagvolden. 1980) prefer choice over no choice. Past researchers have claimed that there is such a thing as ‘the Lure of choice’ (Bown et al. 2003) – that choice in itself is attractive (regardless of a greater outcome), and that humans (and other organisms) are drawn to choice for the sake of choice (Bown et al. 2003).
Whilst popular economical and psychological historical thinking, has indicated self-determination benefits derived from choice for consumers, a number of researchers have identified a limited capacity in humans for processing information (Miller, 1956), (Simon, 1955) and claim that decisions made from large amounts of information can be less than optimal as a result of limiting heuristic processes developed to deal with options for consideration (Wright, 1975), (Tversky, 1972), (Simon, 1955).
Miller’s paper entitled ‘The magical number seven, plus or minus two: some limits on our capacity for processing information’ (Miller, 1956), discussed findings from a number of experiments on information processing capacity limits in humans across several sensory elements, concluding that the capacity to process ‘bits’ of information in humans was at, or around the 6-7 mark, although there is slight variation dependent upon the sense in question. Miller referred to amounts over this as ‘information overload’ (Miller, 1956).
Simon (1955) pointed out that humans are incapable of fully processing overwhelming amounts of information in choices available to us; that some humans (satisficers) adopt heuristics to deal with this. ‘Satisficers’ decide upon an acceptable level across attributes, choose the first option passing that acceptance threshold, and stop searching (satisficing). At the same time, other humans (maximisers), regardless of innate limits, strove to ‘find the best’, seeking to consider all possible options, in an impossible task, before making a choice (maximising) (Simon, 1955).
Others have also labelled human heuristics for handling decision making when managing choice with capacity limits with terms such as simplifying and optimising (Wright, 1975), maximising and satisfising ( Schwartz, 2004), (Simon, 1955) and ‘elimination by aspects’ (Tversky, 1972).
Borrowing the terms ‘maximisers’ and ‘satisficers’ from Simon, Schwartz, in his ‘Tyranny of Choice’ paper categorized thousands of people as either maximisers or satisficers using a ‘maximisation scale’. Schwartz claimed maximisers were less satisfied and more regretful post-decision; that more choice is worse than fewer for them. ‘Maximisers’ are unable to cope with having been unable to evaluate all options and deal with the opportunity costs of ‘loss of alternatives’ (Schwartz, 2004).
The greater the number of options to consider, the more lost opportunities (Tversky, 1972), (Schwartz, 2004). Furthermore, the more options there are, the higher the search costs in terms of time (and arguably money (loss of time)). People ‘adapt’ (get used to what they receive / have / experience) (Schwartz, 2004), meaning that the search costs spent do not always live up to the experiences received from the gains of choosing an option from many.
Iyengar et al, claimed this crosses many domains of decision-making beyond consumer buying behavior, such as job-seeking (Iyengar et al. 2006). Maximisers may achieve more but they feel worse in doing so. (Iyengar et al. 2006).
A growing body of researchers have claimed that having ‘too much choice’ not only produces a negative, demotivating effect, but could even go so far as to lead to reduced performance or feelings of dissatisfaction and regret at having made the wrong choice, post-decision. (Schwartz, 2004), (Iyengar and Lepper, 2000), (Iyengar et al. 2006).
Iyengar and Lepper pointed out that past historical studies advocating more choice had typically considered only small numbers (between 2-6), proposing that this was a significant, overlooked factor and hypothesized that comparing limited (6 items) with extensive ranges (24-30) items would illustrate that too-much-choice has negative, demotivating consequences for consumer buying behaviour and implications for lower quality task performance, with larger numbers at play (Iyengar and Lepper, 2000).
Iyengar and Lepper (2000) undertook a series of 3 experiments, ‘The Jam Study’, ‘The Essay Question’ and ‘The Chocolate Study’.
The Jam Study (Iyengar and Lepper, 2000) – Tasting booths with samples of a limited range (6 items) versus an extensive range (24 items), of exotic jams were set up in a naturalistic retail store. To avoid prior preference issues popular flavors were removed from the choice set. Whilst more people were attracted to the extensive than limited range tasting booth, fewer chose to go on to purchase from the extensive range (3%) when compared with the limited range (30%).
The Essay Question (Iyengar and Lepper, 2000) – Limited (6) and extensive (30) choices were tested in an educational environment to determine impact on motivation and performance quality. Students agreeing to take a non-compulsory essay question were either assigned a list of 6 essay topics or a list of 30 essay topics to choose from. Only 60% of those allocated the 30 essay question topic list submitted completed assignments versus 74% of those allocated 6 topic lists. Of those completing, the quality of the essays of those assigned the 6 topic list was rated as better on average than that of the 30 topic list students despite no prior academic superiority differences recorded beforehand.
The Chocolate Study (Iyengar and Lepper, 2000) – Prior preference skewing was eliminated from this study with those indicating an aversion to chocolate unable to participate, along with prospect subjects expressing a high degree of familiarity with the chocolate brand featured. Participants were assigned to one of three groups (one group subjected to a limited range of chocolates (6), one group to an extensive range (30) , and one control group split into two (half shown 6 and half shown 30 chocolates)). Members of both limited and extensive choice groups could choose and then sample their chosen chocolate from their allocated assortment. Those in the control group were able to choose a chocolate but then received a sample not of their choosing. Participants answered Likert scale questions regarding expectations of satisfaction pre-choice and actual satisfaction post-choice and about their experience of the choosing process.
Findings were that whilst people in the extensive range enjoyed making choices, at the same time they felt both more frustrated and more dissatisfied with the process and the outcome. Participants were found to be significantly less likely to choose chocolates as compensation over money when exposed to an extensive range of chocolates (as too were those in the no-choice control group), when compared to those exposed to limited ranges.
Notably, none of these studies were carried out with subjects with prior preferences – quite the contrary – participants who expressed a preference before a study were excluded (Chocolate study) and popular choices removed from options in advance (Jam study).
In 2003, Chernev found subjects with a prior preference achieved satisfaction from choosing from larger choice sets, but were dissatisfied when they did not have a prior preference.
Chernev claimed that people have an ‘ideal point availability’ (a collection of all of the attributes they are seeking), in advance and that “Those without an idea of what they wish to choose beforehand, face a very complex task indeed of not only searching through an assortment but also of compiling their ‘ideal point availability” (Chernev 2003).
In 2007, Benjamin Schiebehenne, attempted to replicate the results of the 2000 Iyengar and Lepper jam study experiment in an upmarket German store without success (Greifeneder et al. 2010).
(Greifeneder et al. 2010) argued that it is the complexity of the items to consider which caused the ‘too-much-choice’ effect, rather than the number of items themselves – ‘choice complexity’ (Greifeneder et al. 2010). They pointed out that the Iyengar and Lepper Jam Study had involved only uni-dimensional comparison versus multi-dimensional comparisons. Greifeneder et al undertook a study in a semi-replication of the Iyengar and Lepper jam study, this time with images of coloured pens on poster paper. Varying numbers of attributes to consider as well as item numbers themselves were tested. They argued that when the number of attributes to differentiate a choice set on was increased the ‘too-much-choice’ effect occurred, however, when the number of attributes did not differentiate the ‘too much choice’ effect did not occur, regardless of collection size (Greifeneder et al. 2010).
Iyengar and Kamenica in a 2010 study found that people choose the easiest to understand (least complex) option when confronted with large choice sets (Iyengar and Kamenica 2010). In another study involving Iyengar it was found that categorization increased consumer satisfaction levels in subjects without prior preference but had no increased satisfaction effect on those who already knew what they were looking for in advance (Mogilner et al. 2008).
In contrast to ‘too much choice’ advocates, Mochon (2013) claimed ‘not enough choice’ also has negative consequences – that people seek choice when presented with a single item (Mochon 2013); that there is an innate aversion to single item displays. Mochon claimed findings that the ‘single- option aversion’ (Mochon 2013) effect continues beyond the initial search once consumers are exposed to the phenomenon – going on to defer choice beyond the second item if offered further opportunity to seek alternatives.
A number of gaps and conflicts remain across these works.
Firstly, whilst a body of researchers appear to be moving toward a stance of simplicity being key to choosing when a prior preference does not exist, few investigate the impact of online shopping’s dramatic increase in popularity and the simplifying features potentially available via online versus offline consumer experience – and whether this makes choosing easier for consumers – Does too much choice vary between online and offline environments?
Secondly, Iyengar and Lepper studies typically looked at two extremes (limited and extensive ranges), having presented as a major argument that historically low numbers had been compared, but omitted to investigate a middle range assortment sizes in their ‘too much choice’ experiments of 2000.
Thirdly, Mochon speaks of ‘new findings’ in contrast to Iyengar and Lepper study (that people will choose to seek alternatives even if an acceptable option is available) but is this not the earlier identified ‘Lure of Choice’ at play (Bown et al. 2003)?
Fourthly, Iyengar and Lepper studies are with arguably low importance products (chocolates / jam) for which little thought need go into a comparison stage nor presents a particularly enjoyable process. Whilst their study on investment plans constitutes an important life decision it could be argued that this is a ‘grudge’ purchase and results might vary when the product in question is one for which a choosing process is a more enjoyable experience such as a holiday.
Fifth – Iyengar’s studies looked entirely at instances where prior-preference was deterred rather than investigated (Iyengar and Lepper, 2000) – Chernev’s findings are that large choice set results vary with and without prior preference significantly (Chernev 2003).
It is predicted that respondents will choose less choice over more choice in a shop or supermarket environment, but will choose more options to consider simultaneously via a website page because of the increased number of website options now available and designed to simplify ‘choosing’ for consumers (i.e. choice complexity is easier to control for consumers), and consumers are aware of this.
As one of the objectives of this study is to identify an ‘optimum number’ it is impossible to predict an exact optimum comparison figure in advance.
It is predicted that consumers will not defer purchase of an item because there is only a single option available (single-option averson – (Mochon 2013)), if the single option available is one that they have indicated they already have prior preference for.
Research Design & Methodology
The research was of a quantitative design and took the form of an online survey. Respondents were recruited from business contacts, social media contacts, colleagues, family and friends. The nature of the survey was one of anonymous data collection.
Respondents were asked a series of multiple choice questions, one of which contained images of jam assortments, boolean option questions, and one free numerical entry text question.
The majority of questions were designed to achieve objective outcomes, however other questions were included to identify demographics such as age, gender and the online shopping frequency habits of participants.
A major part of the study was designed to partially replicate ‘The Jam Study’ carried out by Iyengar and Lepper in 2000.
To avoid issues with influence from respondents with an aversion to jam, those who answered no to “Do you like jam?” on entry to the survey were not permitted to proceed further.
As jam was chosen as a primary ‘theme’ an initial question was asked to deal with theory of ‘single- option’ aversion (Mochon 2013). Participants were asked to imagine that they were buying jam from a shop (NOT A WEBSITE), and only one option was available (but it was an option that they liked). They were then asked whether they would buy the jam, or not buy the jam because they preferred to compare alternatives?
Continuing with the ‘jam’ theme – One of the questions displayed various images of jam jars in differing volumes (6 jam jars, 12 jam jars, 18 jam jars, 24 jam jars) (as below). 6 and 24 images were chosen as extremes of sample numbers to replicate the Iyengar and Lepper jam study of 2000 which compared a limited range of jams (6) with an extensive range of jams (24). The middle two image sets (12, 18) were included to identify if there was an alternative to these two extremes in a mid range which participants said they preferred.
Participants were asked to imagine that they were in a store (NOT ON A WEBSITE) and asked to choose the number of jam options that they would most prefer to choose from (6, 12, 18, 24)?
6 JAMS TO CHOOSE FROM
(images of jams)
There were a further 3 options with 12, 18 and 24 jam assortment sizes to choose from (see Appendix).
In another question, participants were asked whether they preferred comparing larger numbers of products on a website page versus a shop (offline), and then whether they considered comparing products was easier via a website versus a shop (offline) (multiple choice answers)
Those who answered ‘yes’, they considered comparing products via a website was easier were then provided with a ‘display logic’ question (shown only to them rather than all participants). This question asked them to choose options they considered helped most in comparing products on websites – a number of popular website functions were included in a list of options, such as product filters, product search, categorization, paginated results.
All survey participants were then asked to enter a number between 1-100 (free numerical text entry) to indicate how many products they would feel comfortable comparing on a single website-page.
To gain an understanding of the online-shopping regularity habits of participants a question was also included which asked them to respond how often they bought products online – ranging from ‘never’ to ‘daily’ (multiple choice).
Participants were finally questioned on whether they considered the number of options they would compare when booking a holiday was fewer, more, or the same as if they were buying jam (multiple choice).
Following collection via the online survey, data was analysed via the Qualtrix platform. Whilst the data was anonymous, the initial recruitment sources were pre-planned allowing for some identification of the respondent personas.
This method of data collection (choosing between a number of images of assortments) was selected as it was felt there was a need to carry out a replication of the Iyengar and Lepper ‘Jam Study’. However, this took on board some of the research design of Greifeneder et al 2010 ‘Coloured Pen Study’ in that images rather than real objects were utilised (Greifeneder et al. 2010). Whilst a question focused on the concept of ‘single option aversion’ (Mochon 2013), jam was also used as the product type to ‘theme’ the survey.
Likewise, in the style of Iyengar and Lepper 2010 study those who with prior preferences (an aversion to jam – negative prior preference) regarding one of the main questions (jam images) were not permitted to continue with the study.
This could have been carried out in a retail environment in a fitting replication of the Iyengar and Lepper jam study, whilst simultaneously being carried out on an ecommerce platform for comparison (and carry through to purchase intent behaviour) – however this was deemed unrealistic given time and ethical consideration restrictions.
A number of limitations should be taken into consideration:
Participants numbers – less than half that of Iyengar and Lepper 2010 ‘jam study’. –
Participant profiles – the nature of selected Facebook groups and regular users of social media networks could be deemed as ‘ecommerce savvy’ which may have influenced outcome. – All options presented simultaneously in the question displaying jam images (6,12,18,24) could have an affected. – The study did not carry through to intent to purchase stage (only how comfortable
participants were with choosing from assortment sizes). – The survey design on one question failed to allow respondents to choose more than one option, although it invited participants to do so. – Iyengar / Lepper intentionally used dissimilar exotic jams to avoid prior preference. With the exception of excluding respondents who expressed an aversion to jam, this study did not take further preferences into consideration.
The survey design did not force participant response leading to some variation in response totals. – There was no randomized nature to either the number of images presented in the jam image question (all participants were subjected to a static 6, 12, 18, 24 in ascending order) therefore there was a risk that participants could simply choose the first item which met an acceptable criteria. Randomisation may have gone some way to reduce this risk. – There was no control group in place nor participants receiving different options (other than those receiving display logic output as a result of one question choice. – The survey question regarding ‘single-option aversion’ was a text question which used no images, nor complex multi-dimensionals products and therefore differed to the Daniel Mochon study in this way.
One could argue that choosing jam is not enjoyable when compared, for example, to choosing a holiday, nor is it a major life decision such as that involved in the Iyengar investment study. Whilst one of the questions was designed to address this in part to gain an understanding of whether the phenomenon of ‘too much choice’ could be generalized, the scope must be taken into consideration. – It could be argued that an online survey can replicate a field study in a hand-picked naturalistic environment (as in Iyengar and Lepper, 2000, study) only in part.
Results can be summarised as follows:
Participants were 52% (56) male – 48% (51) female of varying adult age with 94% (103) aged between 25-54.
Of those who were asked whether they liked jam, 84% (106) replied yes and 16% (20) replied no – total 126 (100%).
The survey was terminated for participants responding no, leaving 106 to proceed.
As the study did not force response to all questions, some were unanswered by participants.
When asked whether they would buy an available jam (which they admitted to liking), if it were the only option available or whether they would not buy it because they preferred to always compare to other options, 84% (92) – yes they would buy – 16% (17) – no, they would not buy.
When selecting a preferred number of items to compare from (jam images), in a shop (NOT A WEBSITE) respondents answered as follows:
Total Responses: 100
- 6 items – 51% (51)
- 12 items – 31% (31)
- 18 items – 10% (10)
- 24 items – 8 (8%)
When asked if participants felt more comfortable comparing large amounts of jam on a website over a shop:
- 42% (40) – yes
- 58% (56) – no
56% (54) answered yes, they would consider it easier to compare larger numbers of products on a website page versus a shop, compared with 44% (42) who answered no.
The 54 (56%) of participants who selected yes to choosing on a website being easier than in a shop, when asked to choose website functions making comparing products on a website easier responded as follows:
- Product filters – 42 (78%)
- Product categorisation – 8 (14%)
- Search box – 4 (7%)
- Paginated results – 1 (2%)
- I do not understand the question – 1 (2%)
When asked how many products participants would feel most comfortable viewing simultaneously on a website page (NOT A SHOP) over 50% of the respondents entered numbers above 20, with 21 respondents entering the maximum number of 100. 5 respondents entered numbers which were 6 or below.
Of the 95 respondents who answered the question regarding their online shopping regularity 13% (12) indicated they shopped online less than once a month. 87% (83) shopped online at least monthly with 20% of these (19) 2-3 times per week and 6% (6) daily online shoppers.
66% (62) of respondents who answered the question regarding whether they would prefer to see more options when booking a holiday than they would when buying jam said yes, they would prefer to see more options for holidays than jam. 19% (18) said they would choose the same number of options to consider both and 15% (14) said that they would want to see consider fewer options when booking a holiday.
Full downloadable study results in PDF format
Results Interpretation & Discussion
The study identified a strong tendency to choose the 6 image range size in an offline environment, when choosing jars of jam in an imaginary shop or supermarket.
A pattern emerged, whereby fewer and fewer respondents selected assortment sets as the number of items in them increased. The greater the number of images in the choice set, the less likely participants were to choose that option (as the assortment increased in size, its popularity decreased).
From 100 (100%) participants who answered the ‘how many jams?’ question, 51% (51) chose the most limited range of 6 jams compared with only 8% (8) in the most extensive range of 24 items. In 2nd and 3rd place popularity were 12 and 18 items at 31% (31) and 10% (10) choosers respectively.
However, one must consider context and whether choosing from a large assortment size becomes demotivating across all products with the same ‘optimum’ number of choices remaining, or are there exceptions, for example, when the actual act of choosing is perceived as an enjoyable part of the process? For example – when choosing a holiday.
Findings from this survey indicated that respondents, when questioned, leaned heavily towards comparing larger numbers of items if the subject matter was a holiday (66% – 62) versus jams.
Therefore, it is difficult to generalise jam image optimum choice numbers widely.
The respondents confirmed that in the main they were regular online shoppers as part of the survey.
At least 87% of total participants claimed to shop online at least monthly, indicating that the overall persona of respondents was one with a high degree of familiarity with making online comparisons and choices.
Despite this, findings revealed that participants indicated they found it neither easier, nor more comfortable, to compare large numbers of products online versus offline. However, this was not reflected in the numbers that were entered as optimal range sizes between 1-100 on a website page, which were much higher than the popular option of 6 items offline.
The majority, 78% (42), of those responding positively to finding simultaneous option comparison easier online than offline chose ‘product filters’ as the website function helping most when comparing choices. Given that product filters allow us to sort and filter by attributes it could be argued that this may illustrate an innate tendency to utilize simplifying heuristics such as ‘elimination by aspects’ (Tversky, 1972) and ‘simplifying and optimizing’ (Wright, 1975), or seek ‘ideal point availability’ (Chernev 2003) quickly to overcome ‘choice-complexity (Greifeneder et al. 2010)?
With regards to any ‘single-option aversion’ (Mochon 2013), when asked whether they would not buy jam due to preferring more options to compare, the results were in favour of buying the jam and not continuing to seek alternatives. As respondents were asked to imagine they liked the jam single option available in advance of making their selection, it could be argued choosing yes was always going to be easy given prior preference, (Chernev, 2003).
Findings therefore did not concur with research that even when participants had a prior preference in mind they still had a natural, innate aversion to single options and a desire to search for more choices (Mochon 2013).
However, it is difficult to generalise the results of a survey text question about buying jam and a survey question using images of complex, multi-attribute, high-spec electronic equipment, such as DVD players, which featured in Mochon’s study (Mochon 2013).
Findings Versus Reviewed Literature
The findings did not confirm that people would reject buying a product that they liked simply because there was only one option available. They did not display the ‘single-option aversion’ spoken of by Mochon (Mochon 2013).
The findings did reveal there was a pattern of descending popularity of assortment sizes which correlated with increasing numbers of options. The most popular was 6 items (the most limited range). Interestingly, this concurs with Miller (1956) theories regarding ‘information overload’ occurring at, or around, 6 ‘bits’ of information (Miller, 1956). The findings also correlate with Iyengar and Lepper (2000) view that people would rather choose from a limited range versus an extensive range in a shop or supermarket (Iyengar and Lepper, 2000). However, it contrasts with the view that subjects are drawn naturally initially to more choice. The correlation between increasing assortment sizes and decreasing popularity also indicates a possible perception of greater search costs (Schwartz, 2004), (Tversky, 1972) associated with assortment size – more numbers to compare, the greater the amount of effort required for little gain. However, one could argue that by choosing the smallest number of options participants displayed ‘satisicing’ (Simon, 1956), (Schwartz, 2004.), simply choosing the first option passing an acceptable threshold on the ‘off site’ jam image question.
The findings that participants would be much more prepared to compare larger numbers of items when comparing holidays to jams may further support the search costs theory (Schwartz 2004), (Tversky, 1972) – more effort worth expending upon something more valuable.
Whilst findings did not conclude overall that participants found comparing products online easier or more comfortable than offline, of those who did, their majority choice of ‘product filters’ as the most useful website function may indicate a desire to reduce ‘choice complexity’ (Greifeneder et al. 2010).
Summary of Study Output
The study found that overall participants did not show an aversion to a single option display and that 6 was the assortment size chosen by the most respondents. As the assortment sizes grew larger their was a clear pattern of them becoming less popular as a choice for those choosing. Whilst participants claimed to not find comparing large selections online either more comfortable or easier than offline, further answers to questions did not correspond with this. The nature of ‘too much choice’ phenomenon may be contextual as participants felt that they would want to compare more options for choosing a holiday when compared to choosing jam. The study output may have been influenced in part by the ‘online savvy’ nature of the participants, with the majority indicating that they were regular online shoppers.
Study Aims & Objectives Output
Objective 1 Output
Results were gathered across each of the 4 options for the purposes of this study. The optimum number identified as most preferred when comparing options in a shop (NOT ON A WEBSITE) was 6, however other scope, limitations and risk factors should be taken into consideration with this finding.
The study achieved objective 1.
Objective 2 Output
The study identified that overall participants did not feel that choosing from large selections online would either be more comfortable or easier than choosing offline.
However, when asked to enter a preferred assortment size via a numerical free text between 1 – 100, a large number of respondents chose figures toward the top of this range. This was inconsistent with the popularity of the low ‘optimum’ number of 6 items identified for the offline environment at objective 1.
Whilst the study did not identify a clear preference number online, it indicated a clear difference between comparison set sizes which participants would prefer viewing online versus offline.
The study went some way to achieving objective 2.
Objective 3 Output
The study found that the majority of participants confirmed that they would still buy a product they liked if it was a lone item with nothing to compare. They confirmed that they would not reject the option to buy due to an aversion to single product options – therefore they did not display ‘single-option aversion’ as in Mochon’s study (Mochon 2013).
The study achieved its objective 3.
Recommendations For Future Study
There is potential for further study in a number of related areas. For example, comparing response to optimum online choice numbers across a number of variables such as light versus heavy ecommerce shoppers, gender and age. Of course, further ‘narrowing down’ of the range within which the ‘too much choice tipping point’ occurs may also be feasible.
There is also the potential to compare response to sequential versus simultaneous item display, or paginated results versus infinite scroll output online, building upon Mochon’s theory of ‘single- option aversion’ (Mochon 2013). In an offline retail store further studies aimed at identifying whether products isolated in a retail store sell less – i.e. one product left on a shelf versus surrounded by identical products.
With regards to the prospect of ‘choice complexity’ (Greifeneder et al. 2010) considerations regarding multi-dimensional complex or similar products (e.g. with highly technical specification) versus uni-dimensional comparisons in an online environment may have viable research prospects. This may be of particular use in light of the continuing growth of ecommerce and associated evolving technologies.
Critique Of Own Results
There were a number of areas where the study could be critiqued:
A question asking participants to choose ‘all website functions they believed helped with making item comparison easier on a website’ provided only the option to choose one item on the selection list rather than multiple.
With regards to the exclusion question of ‘Do you like jam’ it became clear some respondents who had initially answered no, had restarted the survey via another browser or device, then answering yes to be involved in the survey, slightly skewing the data (3 responses).
The questions did not force a response meaning some questions were unanswered by respondents.
There is the potential that all items displayed on one page (6,12,18,24) simultaneously could be considered flawed in itself.
The scenario did not reflect a naturalistic field study and did not continue to either a real or imagined intent to purchase, covering only preference of assortment sizes – no ‘real’ choices with potential consequences were made.
The survey was intentionally short due to concern over limited attention span of participants compensated with goodwill only – with risk the survey did not gather enough crucial information.
Further research design piloting may well have helped to identify a better ideal length (number of questions) for this study. With further work, permissions and ethics considerations this study could be continued using a live ecommerce environment.
Appendices & Results
# Answer Response %
1 19 – 24 2 2% 2 25 – 34 32 29% 3 35 – 44 43 39% 4 45 – 54 28 26% 5 55 – 64 2 2% 6 65+ 2 2% Total 109 100%
# Answer Response %
1 Male 56 52% 2 Female 51 48% Total 107 100%
‘NON-JAM-LIKERS’ V ‘JAM-LIKERS’ Question: Do you like jam?
# Answer Response %
1 Yes 106 84% 2 No 20 16% Total 126 100%
PARTICIPANTS WHO WOULD BUY A SINGLE OPTION
Question: Imagine you go to buy jam in the local supermarket and there is only one jam option available on display, but it is jam you like which is available.
Would you buy the jam or would you not buy the jam as you prefer more options to compare?
# Answer Response %
1 I would buy it 92 84% 2 I would not buy it because I
prefer to see more options
Total 109 100%
HOW MANY JAMS?
WHEN ASKED HOW MANY JAM OPTIONS PARTICIPANTS WOULD PREFER TO CHOOSE FROM?
# Answer Response %
1 6 49 51% 2 12 29 30% 3 18 10 10% 4 24 8 8% Total 96 100%
QUESTION: Do you think that you would feel more comfortable comparing larger numbers of options on a website page, for jam, than you would if you were visiting a shop / supermarket?
# Answer Response %
1 Yes 40 42% 2 No 56 58% Total 96 100%
QUESTION: Do you consider it easier to compare larger numbers of products on a website page than in a shop?
# Answer Response %
1 Yes 54 56% 2 No 42 44% Total 96 100%
QUESTION: Check all the items (website functions) that you would use to make comparing larger numbers of products on a website easier than comparing larger numbers of products in a shop
# Answer Response %
2 Product Filters (e.g. filter by size, colour, brand, etc) 42 78%
1 Product categorisation / Departments within the website 6 11% 3 Search box 4 7% 5 I do not understand the question 1 2% 4 Paginated results (page 1, page 2, page 3, etc) 1 2% Total 54 100%
QUESTION: How many products per page (jam or other products) do you feel most comfortable viewing at the same time on a website page (NOT A SUPERMARKET OR SHOP) – (i.e. how many results per page feels comfortable to you?
Please enter a number between 1 and 100 in the box below
QUESTION: How often do you shop online (for any products)?
How often do you shop online? 1
# Answer Response %
1 Never 2 2% 2 Less than Once a Month 10 11% 3 Once a Month 15 16% 4 2-3 Times a Month 24 25% 5 Once a Week 19 20% 6 2-3 Times a Week 19 20% 7 Daily 6 6% Total 95 100%
QUESTION: Finally – Imagine you are looking to book a holiday?
Do you think that you would prefer to see more or fewer holidays on display, either in a shop (travel agent) (e.g. on the travel agent window), or on a website page, at the same time, than you would choose for the number of jams you would prefer to see displayed?
Imagine you are looking to book holiday 1
# Answer Response %
1 Yes – more options displayed for holidays than I would
choose for jam
2 The same number of options for holidays displayed as I
would choose for jam
3 No – fewer options displayed for holidays than I would
choose for jam
Total 94 100%
Baddeley, Alan (1994) – The Magical Number Seven: Still Magic After All These Years? – Psychological Review, ISSN 0033-295X, 04/1994, Volume 101, Issue 2, p. 353
Bown, Read, Summers (2003) – The Lure Of Choice – Journal of Behavioural Decision Making, ISSN 0894-3257, 10/2003, Volume 16, Issue 4, pp. 297 – 308
Catania, A.C. (1975). Freedom and knowledge: an experimental analysis of preference in pigeons. Journal of the Experimental Analysis of Behaviour, 24, 89-106.
Catania, A.C. & Sagvolden. T (1980) – Preference for free choice over forced choice in pigeons. Journal of the Experimental Analysis of Behaviour, 34, 77-86.
Chernev, Alexander (2003) – When More Is Less and Less Is More: The Role of Ideal Point Availability and Assortment in consumer Choice, Journal of Consumer Research, Vol. 30, No. 2 (September 2003), pp. 170-183
Dhar Ravi (1996) – The Effect On Decision Strategy On Deciding To Defer Choice, – Journal of Behavioral Decision Making, ISSN 0894-3257, 12/1996, Volume 9, Issue 4, pp. 265 – 281
Dhar, Ravi (1997) – Consumer Preference For A No-Choice Option; Journal of Consumer Research, ISSN 0093-5301, 09/1997, Volume 24, Issue 2, pp. 215 – 231
Dhar, Ravi (1997) – Context and Task Effects On Choice Deferral – Marketing Letters, ISSN 0923-0645, 01/1997, Volume 8, Issue 1, pp. 119 – 130
Dhar, Ravi and Simonson, Itamar (2003) – The Effect Of Forced Choice On Choice:- Journal of Marketing Research, ISSN 0022-2437, 05/2003, Volume 40, Issue 2, pp. 146 – 160
Greifender, Rainer; Scheibehenne, Benjamin; Kleber, Nina (2010) – Less may be more when choosing is difficult: Choice complexity and too much choice – Acta Psychologica (Science Direct), Volume 133, Issue 1, January 2010, P45-50
Iyengar Sheena S, Kamenica, Emir (2010) – Choice proliferation, simplicity seeking, and asset allocation, Journal of public economics, ISSN 0047-2727, 2010, Journal of Public Economics, Volume 94,Issue 7/8, pp. 530 – 539
Iyengar, Leppar, 2000 – When choice is demotivating: Can one desire too much of a good thing? Journal of personality and social psychology [0022-3514] Iyengar, Sheena yr: 2000 vol:79 iss:6 pg:995 -1006
Iyengar, Schwartz, Wells (2006) – Doing Better But Feeling Worse – Psychological Science February 2006 vol. 17 no. 2 143-150
Miller, George, A (1956) – The magical number seven, plus or minus two: some limits on our capacity for processing information. Psychological Review, ISSN 0033-295X, 03/1956, Volume 63, Issue 2, pp. 81 – 97
Mochon, Daniel (2013) – Single Option Aversion: Journal of Consumer Research, ISSN 0093-5301, 10/2013, Volume 40, Issue 3, pp. 555 – 566
Mogilner, Cassie; Runick, Tamar; Iyengar, Sheena S (2008) – The Mere Categorization Effect: How the Presence of Categories Increases Choosers’ Perceptions of Assortment Variety and Outcome Satisfaction: Journal of Consumer Research, ISSN 0093-5301, 08/2008, Volume 35, Issue 2, pp. 202 – 215
Oppewal, Harmen and Koelemeijer, Kitty (2005) – More Choice is Better: Effects of assortment size and composition on assortment evaluation – International Journal of Research in Marketing, ISSN 0167-8116, 2005, Volume 22, Issue 1, pp. 45 – 60
Scheibehenne, Benjamin (2010) – Can There Ever Be Too Many Options? A Meta‐Analytic Review of Choice Overload, The Journal of consumer research [0093-5301] yr: 2010 vol:37 iss:3 pg:409 -425
Barry Schwartz (2004) – The Paradox Of Choice: New York: HarperCollins, 2004
Barry Schwartz (2004) – The Tyranny Of Choice, – Scientific American, 290, 70-76 – 2004
Simon, H.A. (1955). A Behavioural Model Of Rational Choice, Quarterly Journal of Economics, 59, 99-118.
Spassova, Isen (2013) – Positive affect moderates the impact of assortment size on choice satisfaction (Spassova, Isen) – Journal of Retailing [0022-4359] Spassova, Gerri yr:2013 vol:89 iss:4 pg:397 -408
Suzuki, S (1999) – Selection of forced and free-choice by monkeys (Macaca Fascicularis). Perceptual and Motor Skills, 88, 242-250.
Tversky, Amos (1972) – Elimination by aspects: A theory of choice. Psychological Review, volume 79, issue 4, 281-299.
Voss, S.C., & Homzie, M.J. (1970). Choice as a value. Psychological Reports, 26, 912-914.
Wright, Peter (1975) – Consumer Choice Strategies: Simplifying vs. optimizing: – Journal of marketing research, ISSN 0022-2437, 1975, Volume 12, Issue 1, pp. 60 – 67
Zuckerman – Porac – Lathin – Smith – Deci – 1978 – On The Importance Of Self-Determination For Intrinsically-Motivated Behavior – Personality and Social Psychology Bulletin, ISSN 0146-1672, 07/1978, Volume 4, Issue 3, pp. 443 – 446