When and why we forget to buy

Available online at www.sciencedirect.com

ScienceDirect Journal of Consumer Psychology 26, 3 (2016) 363 – 380

Research Article

When and why we forget to buy☆ Daniel Fernandes a,⁎, Stefano Puntoni b , Stijn M.J. van Osselaer c , Elizabeth Cowley d a

Católica-Lisbon School of Business and Economics, Universidade Católica Portuguesa, Portugal b Rotterdam School of Management, Erasmus University, The Netherlands c Samuel Curtis Johnson Graduate School of Management, Cornell University, USA d University of Sydney Business School, Australia Accepted by Cornelia Pechmann, Editor; Associate Editor, Maureen (Mimi) Morrin Received 24 January 2013; received in revised form 23 June 2015; accepted 26 June 2015 Available online 3 July 2015

Abstract We examine consumers' forgetting to buy items they intended to buy. We show that the propensity to forget depends on the types of items consumers intend to purchase and the way consumers shop. Consumers may shop using a memory-based search by recalling their planned purchases from memory and directly searching for the products. For example, consumers may use the search function at an online store. Alternatively, consumers may use a stimulus-based search by systematically moving through a store, visually scanning the inventory and selecting the required items as they are encountered. Using an online shopping paradigm, we show that consumers are more likely to forget the items they infrequently buy when using the memory-based search, but not when using the stimulus-based search. In fact, when using the stimulus-based search, consumers are sometimes even better able to remember the items they infrequently (vs. frequently) buy. Moreover, consumers fail to take these factors into account when predicting their memory. As a result, they do not take appropriate actions to prevent forgetting (e.g., using a shopping list). © 2015 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved. Keywords: Memory; Metamemory; Consideration sets; Shopping lists; Memory-based search; Stimulus-based search

Introduction Forgetting to buy is both common and annoying. As consumers, we have all forgotten to buy an ingredient necessary for a meal or an item our significant other asked us to purchase just minutes before we entered the grocery store. For consumers, the consequences of forgetting include having to return to the ☆ The financial assistances of the Erasmus Research Institute of Management (ERIM) and the Católica-Lisbon Research Unit on Business and Economics (CUBE) are gratefully acknowledged. This paper is based on an essay of the first author's dissertation. The authors thank the members of the Consumer Financial Decision Making lab group at the University of Colorado for comments and discussion of earlier versions of the paper. The authors also thank Annette Bartels, Leonor Machado and Patricia Neto for help with data collection. ⁎ Corresponding author. E-mail addresses: [email protected] (D. Fernandes), [email protected] (S. Puntoni), [email protected] (S.M.J. van Osselaer), [email protected] (E. Cowley).

store or to re-organize meal plans. For companies, consumer forgetting results in missed sales. Forgetting to buy is a problem for both routine purchases, such as forgetting to buy milk when it is needed for breakfast, and for less routine occasions, such as forgetting to buy asparagus when it is needed for a special meal. Grocery shopping, either in brick and mortar or online stores, represents one of the largest household expenditures (Food Marketing Institute, 2012). Despite its importance, no research has investigated factors that cause consumer forgetting while shopping. Descriptive and correlational results indicate that forgetting is common, and can be substantially reduced by the simple use of shopping lists (Block & Morwitz, 1999; Hui, Inman, Huang, & Suher, 2013). We address this gap by looking at when and why shoppers forget to buy. We argue that people are not equally likely to forget to buy in all situations or for all products. We distinguish between two different shopping strategies and assess how the type of search in each shopping strategy impacts the likelihood of forgetting items that are

http://dx.doi.org/10.1016/j.jcps.2015.06.012 1057-7408/© 2015 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.

364

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

either high or low in purchase frequency, such as milk and asparagus in the examples above. In order to understand consumer forgetting, we need to consider consumers' predictions of their memory performance. If consumers can accurately predict forgetting they can make informed decisions about what, if any, preventative measures should be employed to reduce the likelihood of forgetting (e.g., write a shopping list or use the best search strategy depending on the type of items they need). Therefore, we investigate whether consumers' memory predictions accurately reflect memory performance and, in particular, the influence of search strategy and frequency of purchase of items on actual memory. Below, we present our hypotheses of how shopping strategy and the frequency of item purchase may affect forgetting. Similar to the distinction concerning choices (Lynch & Srull, 1982), we distinguish between a stimulus-based search strategy, where consumers systematically scan items in the store to find the products they need, and a memory-based search strategy in which consumers try to recall items from memory to guide search directly to the item's location. Across a series of studies, we find that search strategy interacts with frequency of item purchase, such that items that are infrequently purchased are forgotten more often than items that are frequently purchased when consumers are using the memory-based search strategy. This effect for item type is not observed when using the stimulus-based search strategy. Importantly, consumers do not anticipate the shopping strategy by purchase frequency interaction in their memory predictions. Theoretical background Grocery shopping: a challenge for memory Consumers have difficulty remembering previously identified grocery needs (Bettman, 1979). In fact, on average consumers fail to buy about 30% of items they intended to buy (Hui, Huang, Suher, & Inman, 2013). For example, during the 2010 Christmas season, 44% of consumers forgot to buy thank you notes, 37% forgot to buy holiday cards or letters, 36% forgot to buy batteries (for Christmas gifts), and 25% forgot to buy wrapping paper (New York Times, 2011). These unfulfilled purchases occur less often when a memory aid, such as a shopping list, is used (Block & Morwitz, 1999). However, many shoppers do not use shopping lists. Descriptive studies from several countries show that only about half of shoppers use lists (Thomas & Garland, 2004). In sum, research in consumer behavior shows that memory while shopping is highly fallible (Bettman, 1979) and that shoppers need a shopping list to remember (Block & Morwitz, 1999). However, many shoppers do not use a shopping list to remember what they need to buy. Thus, forgetting is a problem for shoppers. In the next section, we build the hypotheses about when and why forgetting to buy is most common. Two ways of shopping: memory-based and stimulus-based search Consumers may choose to shop for groceries by recalling their planned purchases from memory, locating the recalled products, and putting them in their shopping cart. Consider a

shopper using this strategy when shopping for cheddar cheese. In the store, the consumer would first try to recall the items s/he was planning to buy, hopefully retrieving cheddar cheese. S/he would then go to the Dairy Products section and pick up the cheese. We call this memory-based search. Alternatively, shoppers may choose to methodically move through each of the sections of the store visually scanning the items to ensure all of the required products are selected (e.g., browsing all product categories in an online store or walking up and down the aisles in a brick and mortar store). In this case, a shopper may see the cheddar cheese while moving through the Dairy section, and realize that it is needed. We call this a stimulus-based search. Thus, shoppers may retrieve a needed item from memory and go straight to the item location or browse through the categories in search of the products they need. Previous research has made a distinction between memorybased and stimulus-based choices (Lynch & Srull, 1982; Rottenstreich, Sood, & Brenner, 2007; Sanbonmatsu & Fazio, 1990). In memory-based choices, consumers make a choice by recalling the options, whereas in stimulus-based choices, consumers make a choice while being exposed to the options. These two different approaches to choice influence what options are considered and chosen. We posit that the distinction between memory-based and stimulus-based can be extended to how consumers search for a planned purchase. That is, after consumers have decided to purchase an item, the strategy consumers use to search for the items they need (i.e., memorybased vs. stimulus-based search) can influence whether they end up buying that item. It is important to highlight that the memory-based versus stimulus-based distinction for search strategies is not an absolute distinction but more a matter of degree. For example, even though the stimulus-based search strategy relies more on stimulus information than on memory, one has to match the products one is scanning with memory about whether the product is needed. Similarly, in many situations in which search is predominantly memory-based, consumers may still rely quite heavily on stimulus-based visual information to physically locate products after retrieving a to-be-bought product from memory. Although prior research has not addressed these search strategies per se, studies documenting consumers' paths through stores provide support for their existence (Hui, Bradlow, & Fader, 2009; Hui, Huang, et al., 2013; Hui, Inman, et al., 2013). Prior research also suggests that although both strategies (and a mix of the two strategies) are common, the majority of consumers do not use a pure stimulus-based strategy (Bell, Corsten, & Knox, 2011; Hui et al., 2009; Hui, Huang, et al., 2013; Hui, Inman, et al., 2013; Inman, Winer, & Ferraro, 2009; Stilley, Inman, & Wakefield, 2010) despite the fact that walking the aisles helps memory (Gilbride, Inman, & Stilley, 2015; Hui, Huang, et al., 2013) as in-store shelf facings benefit memory, consideration, and choice (Chandon, Hutchinson, Bradlow, & Young, 2009). In addition, walking the aisles is often stimulated by retailers working hard to motivate shoppers to browse their assortment. Examples include scattering products that are regularly bought (e.g., milk, eggs) around the store (Granbois, 1968; Iyer, 1989), forcing customers to walk the entire store (e.g., Ikea, Hollister),

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

365

and using in-store promotions and mobile applications (e.g., Foursquare, Shopkick) that motivate shoppers to visit more store areas (Hui, Inman, et al., 2013). In sum, prior research supports the existence of memory-based and stimulus-based search strategies. Given the fact that consumers often do not use physical (vs. mental) shopping lists, we assume that they rely on memory-based and/or stimulus-based search to remember the items on their mental lists.

recognition-based strategy and memory-based search a recall-based strategy. Analogous to the word-frequency paradox findings, we propose that shoppers are better able to remember items that are frequently purchased compared to items that are infrequently purchased when using the memory-based search strategy, whereas the frequency of purchase advantage disappears, and may even reverse, when using the stimulus-based search strategy.

Frequently purchased versus infrequently purchased items

Miscalibration in memory predictions

There are differences between the types of products shoppers intend to buy which we believe will affect whether the items are easily remembered when shopping. One such difference is the frequency with which shoppers purchase or consume products. Shoppers may intend to buy items they purchase or consume every week or items they purchase or consume only occasionally. Therefore, a consumer's mental shopping list may include items which will replenish the stock of frequently purchased items consumed in the last few days and/or items which are rarely consumed. Previous consumer research has not examined the effects of frequency of purchase of a product category, but has considered the frequency of purchasing a particular brand on memory, consideration, and choice. Some researchers have found that frequently purchased brands are more accessible and thus more likely to be considered (e.g., Hauser & Wernerfelt, 1990; Janiszewski, Kuo & Tavassoli, 2013; Nedungadi, 1990; Pechmann & Stewart, 1990). Others researchers have found that infrequently purchased brands, such as those from low market share brands or lower price alternatives, are at least as likely to be considered, and by implication remembered, as frequently purchased brands (e.g., Bemmaor & Mouchoux, 1991; Chandon et al., 2009; Diehl, Kornish, & Lynch, 2003; Mogilner, Shiv, & Iyengar, 2013). These seemingly contradictory results may be reconciled by considering how people remember information. The wordfrequency paradox posits that common words are better recalled than rare words, while rare words are more easily recognized than common words (e.g., Dobbins, Kroll, Yonelinas, & Liu, 1998; Gregg, 1976; Hemmer & Criss, 2013; Kintsch, 1967; Reder et al., 2000). The explanation for the paradox is that common words have more associations with other words than rare words which makes common words easier to retrieve and, therefore, more likely to be remembered when using a recall strategy (e.g., Gregg, 1976; Hemmer & Criss, 2013; Kintsch, 1967). On the contrary, because rare words stand out from other words, they are easier to discriminate and, therefore, easier to recognize (e.g., Dobbins et al., 1998; Hemmer & Criss, 2013; Reder et al., 2000). There are striking similarities between shopping search strategies and recall versus recognition tasks. For instance, a stimulus-based search strategy and recognition tasks require one to discriminate target items from among a set of distracting items. Conversely, a memory-based search strategy and recall tasks require one to actively retrieve an item from memory. Thus, we propose that stimulus-based search is essentially a

If people can perfectly predict how well they will remember grocery items they need when shopping, then they can make informed judgments about whether they need to take preventative measures (e.g., rehearse the items on the mental list, write a physical shopping list, or use the best search strategy to remember to buy a certain type of item). If that is the case, the interaction between frequency of purchase and search strategy on memory performance described earlier is less consequential as consumers can decide whether to prevent its effect. However, if memory predictions are poorly calibrated, the interaction between frequency of item purchase and search strategy on memory performance predicted above will result in under- (or possibly even over-) prepared shoppers. The question thus becomes: are people sensitive to the effects of frequency of item purchase and search strategy outlined above? That is, are consumers' memory predictions well-calibrated? Memory predictions have been examined under the label of metamemory and refer to “people's knowledge of, monitoring of, and control of their own learning and memory processes” (Dunlosky & Bjork, 2008, p. 11). Memory predictions depend heavily on the accessibility of items when the predictions are made and fail to take into account all the factors that influence memory after a delay (e.g., Benjamin, Bjork, & Schwartz, 1998; Koriat, Bjork, Sheffer, & Bar, 2004; Kornell & Bjork, 2009; Kornell, Rhodes, Castel, & Tauber, 2011; Schwartz, 1994). People estimate their memory performance based on how easily they can access the items at the time of memory predictions, not on the conditions in the retrieval environment. Thus, we expect that people do not spontaneously anticipate that memory performance depends on the interaction between frequency of item purchase and search strategy. Summary of predictions and studies To summarize our predictions, we hypothesize an interaction between search strategy and frequency of item purchase on consumers' memory performance for grocery products. Specifically, when people shop using a memory-based search strategy, they will be more likely to forget products they infrequently buy (compared to products they frequently buy). The difference in the likelihood of forgetting between items low and high in frequency of purchase will not be observed in the case of a stimulus-based search strategy and may even reverse. We predict that consumers' memory predictions will not spontaneously be sensitive to the interaction between search

366

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

strategy and frequency of item purchase on actual memory. Thus, we expect the interaction to occur in memory performance (how many items are remembered when shopping), but not in memory predictions (how many items they predict they will remember to buy). In Studies 1a and 1b, we manipulated search strategy on an online grocery store website. One third of participants were instructed to buy the items on the website using the search bar to find the products they needed (memory-based search), another third of participants were instructed to browse the categories of the store to find the products they needed (stimulus-based search), and the remaining third were not given explicit search instructions (control). Before shopping, we gave participants a list of 10 grocery items and asked them how many of the items they thought they would remember to buy if the shopping trip occurred 5 min later. Then, participants completed a series of unrelated tasks for about 8 min on average. After the delay, participants shopped at the online store. We measured how many of the 10 items they bought. We observed that unpredicted by participants, memory was better for frequently purchased than for infrequently purchased items in the memory-based search condition, but not in the stimulus-based search condition. Study 1b replicated these results with a 10-minute delay constant for all participants and a different participant sample. In Study 2, we provide process evidence for the interaction between search strategy and frequency of item purchase by showing that frequently purchased grocery items are remembered better than infrequently purchased grocery items in a standard recall task whereas this advantage disappears in a standard recognition task. In Studies 3a and 3b, we explore lay theories about memory to assess whether consumers are able to accurately predict their memory performance when asked explicitly about the effects of frequency of item purchase and search strategy. One possibility is that consumers are naive; another is that they can correctly introspect on the effects of purchase frequency and search strategy when the independent variables are highlighted for them, but do not spontaneously consider these beliefs about forgetting. In line with the latter, consumers in Studies 3a and 3b did accurately identify the conditions under which they are most likely to forget when we explicitly asked them about how our independent variables would impact forgetting to buy. Study 1a: shopping for groceries at Coles In Study 1a, we used an existing online grocery store to test our core predictions (1) that consumers are less likely to forget frequently-purchased grocery items than infrequently-purchased grocery items when they rely on a memory-based search strategy but not when they rely on a stimulus-based search strategy and (2) that consumers do not spontaneously predict this interaction effect before shopping. Grocery retailers' websites often allow consumers to shop in two different ways: the website provides a search function where consumers can type the products they are searching for (memory-based search strategy) and the website also provides drop-down or side-bar menus with categories and

items that allow consumers to browse (stimulus-based search strategy). Thus, in Study 1a, we asked participants to shop for a list of products using an actual online grocery retailer website and instructed them to use memory-based search, stimulus-based search, or offered no explicit search instructions. We acknowledge that the memory-based and the stimulus-based search are not orthogonal as the use of the drop-down menu could still involve a bit of memory-based search. However, we expected participants to employ the stimulus-based search strategy more often when using the drop-down menu than when using the search bar, whereas the opposite is true for the memory-based search strategy. Method Participants and design One hundred forty-five students at a major Southwest European university participated in the study in return for course credit (Mage = 22.71, SD = 1.51; 42% female). The key independent variables were frequency of item purchase (frequent vs. infrequent) and search strategy (memory-based vs. stimulus-based vs. control), both variables were manipulated between-participants. For each participant, we measured their pre-shopping prediction of how many items they would remember to buy and, after a delay, we measured the actual number of items they remembered to buy in an online grocery store. Procedure Participants were informed that the study aim was to test whether people are able to remember to buy the products they need to buy. Participants in the memory-based search strategy condition were told that “5 minutes from now, we'll ask you to enter an online grocery store to buy the 10 fruits & vegetables below by typing their names into the store's search engine”. Participants in the stimulus-based search strategy condition were told that “5 minutes from now, we'll ask you to enter an online grocery store to buy the 10 fruits & vegetables below by browsing the categories in the online store”. And participants in the control condition were told that “5 minutes from now, we'll ask you to enter an online grocery store to buy the 10 fruits & vegetables below”. All participants were then provided a list of 10 items. For about half of participants, the list included frequently purchased items (apple, banana, broccoli, lettuce, papaya, potato, pumpkin, strawberry, tomato, and watermelon). For the other half, the list included infrequently purchased items (avocado, beetroot, celery, coconut, fig, mushroom, passion fruit, pineapple, plums, and rhubarb). We selected these items by examining what products people most frequently bought and infrequently bought in the fruits & vegetables category of the Coles online store. Immediately after the presentation of the 10 items, participants predicted how many of the items they would remember to buy when shopping 5 min later. Specifically, participants in the memory-based search condition answered the question “how many of the 10 products above do you think you will remember and be able to type on the search engine of the online grocery store 5 minutes from now?” Participants in the stimulus-based search condition

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

answered the question “how many of the 10 products above do you think you will remember and be able to identify in the online grocery store 5 minutes from now?” Participants in the control condition answered the question “how many of the 10 products above do you think you will remember to buy in the online grocery store 5 minutes from now?” After participants indicated the number of products they thought they would remember, a time interval followed in which they completed a series of unrelated tasks. Participants, on average, took 8 min and 9 s to complete the unrelated tasks (SD = 2.29). After completing the unrelated tasks, participants were asked to find the products in the online store (http://shop.coles. com.au) operated by Coles, a leading Australian retailer (see Fig. 1 for a screen capture of a Coles.com.au page instructing consumers on how to shop). Those in the memory-based condition were given instructions to find the products by only using the search engine of the store and typing the products’ names into the search bar (see Fig. 2). They were also given explicit instructions not to browse the website in search for the products. Those in the stimulus-based search condition were given instructions to find the products by only browsing the website in the Shop By Category option and looking for the products inside the Fruits & Vegetables category (see Fig. 3, note that all the items are listed in the store cuing participants for the products they need to buy). They were also given explicit instructions not to use the search engine of the store to find the products. Those in the control condition were simply told to find the products. Memory performance was measured by the number of items purchased from the list. Next, participants answered manipulation check questions for frequency of purchase and shopping strategies. For the manipulation check of frequency of purchase, participants were asked to indicate their familiarity with the items (− 5 = very unfamiliar; 5 = very familiar), how frequently they consume the items (− 5 = very infrequently; 5 = very frequently) and

367

how frequently they purchase the items (− 5 = very infrequently; 5 = very frequently). For the manipulation check of shopping strategy, participants were asked to indicate how they went about finding the products among the following options: 1 = I only used the search function; 2 = I mostly used the search function but also browsed the category a little; 3 = I used the search function and browsed the category about the same; 4 = I mostly browsed the category but also used the search function a little; and 5 = I only browsed the category. We report additional details of the procedure of our studies in the methodological details Appendix A. Results For all experiments, all measures, conditions, and data exclusions were reported. In all studies, we inspected the data for outliers (cases in which measures were lower or higher than 3 standard deviations from the mean). No outliers were detected in this study. In addition, in all studies, sample sizes were determined by attempting to have at least 20 participants per condition. Manipulation checks We averaged the three manipulation check items for purchase frequency (α = .76; M = 1.19; SD = 2.43). As expected, participants in the infrequently purchased items conditions were less familiar with the items they had to remember, consumed these products less frequently, and bought these products less frequently (M = − 0.05, SD = 2.27) than those in the frequently purchased items conditions (M = 2.41, SD = 1.92; F(1, 143) = 49.53, p b .01). Regarding shopping strategy, participants who received instructions to use a stimulus-based search strategy reported that they browsed the categories on the website more (M = 4.32, SD = 1.37) than those using memory-based search strategy (M = 1.33, SD = 0.95; F(1, 142) = 119.05, p b .01)

Fig. 1. Shopping at Coles (study 1). Screen capture of a page of the Coles online store in which shoppers are explained how to shop.

368

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Fig. 2. Memory-based search strategy condition (study 1). Screen capture of the Coles online store where participants used the memory-based search strategy.

and those in the control condition (M = 2.27, SD = 1.57; F(1, 142) = 57.61, p b .01). Those in the memory-based search strategy browsed the categories less than those in the control condition (F(1, 142) = 12.40, p b .01). Hypothesis tests We analyzed the data using a repeated measures ANOVA on memory predictions and performance where the frequency of item purchase (frequently vs. infrequently purchased) and shopping strategy (memory-based search vs. stimulus-based

search vs. control) were entered as between-participants factors. Following the convention in the metamemory literature (e.g., Benjamin et al., 1998; Koriat et al., 2004; Kornell et al., 2011), we entered predicted and actual memory performance as repeated measures. That is, the number of items participants predicted they would remember (memory predictions) versus the number of items they actually remembered (memory performance) was entered as a within-participants factor. We found a three-way interaction between frequency of item purchase, shopping strategy and memory predictions vs. performance (F(2, 139) =

Fig. 3. Stimulus-based search strategy condition (study 1). Screen capture of the Coles online store where participants used the stimulus-based search strategy.

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

2.87, p = .06, see Fig. 4). We decomposed this three-way interaction by looking at the two-way interaction between frequency of item purchase and shopping strategy for each of the two dependent measures (predicted and actual memory performance) separately. For actual memory performance, we found significant main effects of shopping strategy (Mmemory-based = 6.19, SD = 2.54, vs. Mcontrol = 6.21, SD = 2.74, vs. Mstimulus-based = 7.26, SD = 2.10; F(2, 139) = 3.21, p = .04) and of frequency of item purchase (Mfreq = 7.19, SD = 2.14, vs. Minfreq = 5.88, SD = 2.70; F(1, 139) = 10.72, p b .01), and, crucially, a significant interaction between these two factors (F(2, 139) = 2.99, p = .05). As expected, participants remembered to buy frequently purchased items significantly better than infrequently purchased items in the memory-based search strategy condition (Mfreq = 7.12, SD = 2.11, vs. Minfreq = 5.17, SD = 2.62; F(1, 139) = 8.09, p b .01) or when not provided with search instructions (Mfreq = 7.24, SD = 2.18, vs. Minfreq = 5.23, SD = 2.89; F(1, 139) = 9.17, p b .01), but not when using the stimulus-based search strategy (Mfreq = 7.22, SD = 2.23, vs. Minfreq = 7.30, SD = 2.01; p = .90). As hypothesized, fewer items were remembered in the infrequently (vs. frequently) purchased item condition in memory-based search strategy condition, but not in the stimulus-based search strategy condition. For memory predictions, we only found that memory was predicted to be higher for frequently purchased items than for infrequently purchased items (Mfreq = 6.54, SD = 1.84 vs. Minfreq = 5.49, SD = 1.96; F(1, 139) = 10.80, p b .01). There was no effect of search strategy (p = .43), and no interaction between these two factors (p = .88). Thus, participants did not anticipate the interaction effect of shopping strategy and item purchase frequency on actual memory performance. Study 1a confirmed our predictions (1) that consumers are less likely to forget frequently-purchased grocery items than infrequently-purchased grocery items when they rely on a memory-based search strategy, but not when they rely on a stimulus-based search strategy and (2) that consumers do not spontaneously predict this interaction effect before shopping. In study 1b, we replicated Study 1a with a larger sample of US-based participants and a slightly longer, fixed delay between memory prediction and actual memory performance. Number 9. of items 8.

*

Study 1b: shopping for groceries at Coles II In study 1b, we replicated study 1a with two changes to the procedure and one change to the materials. First, instead of European undergraduates, participants in Study 1b were US-based workers recruited via Amazon's Mechanical Turk (MTurk), who may be more experienced with grocery shopping than undergraduate students. In addition, we increased the time interval between memory prediction and the shopping task to 10 min and used a larger sample to explore whether the effect of frequency of purchase in the memory-based condition could not only disappear but could actually reverse in the stimulus-based condition. A small change to the materials was to use oranges instead of papaya as one of the frequently purchased items as U.S. shoppers are more likely to purchase oranges frequently. Method Participants and design Four hundred sixty-three MTurk workers in the U.S. participated in the study in exchange for US $0.50. Fifty-four participants did not pass an instruction manipulation check at the beginning of the study. In addition, memory performance for two of the participants was lower than 3 standard deviations from the mean and were, therefore, excluded. (See methodological details Appendix B for analyses that include these participants.) This left us with 407 participants (Mage = 37.83, SD = 12.65; 62% female). When all 463 responses are included, the pattern of results remains the same. As in Study 1a, the key independent variables were frequency of item purchase (frequent vs. infrequent) and search strategy (memory-based vs. stimulus-based vs. control), both variables were manipulated between-participants. We again measured pre-shopping memory predictions of how many items participants would remember to buy in the shopping task and memory performance of how many items they could actually remember to buy in the shopping task. Procedure We employed an instruction manipulation check (Oppenheimer, Meyvis, & Davidenko, 2009) at the beginning of the study given

* 7.22 7.30

7.24

7.12

7. 6.

369

6.70

6.60

6.32

5.87 5.23

5.17

5.38

5.22

5.

Frequently purchased items

4. Infrequently purchased items

3. 2. 1. . Memory-based

Control

Stimulus-based Memory-based

Memory Performance

Control

Stimulus-based

Memory Predictions

Fig. 4. Study 1a results. Memory predictions and performance as a function of type of items and of search strategy. Note: *significant pairwise tests reported in the text.

370

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

that participants were required to strictly follow the instructions of search strategies. After the instruction manipulation check, the procedure was the same as in Study 1a except that the time interval between the item presentation phase and the online shopping task was 10 min rather than 5 min. During that time interval, participants read the first 2089 words of the story “The Case of the Velvet Claws” by Erle Gardner. Participants were asked to read the passage carefully. They were informed that they had exactly 10 min to read the story. A timer was displayed on the screen. After 10 min, the study automatically moved forward. Results Manipulation checks We averaged the three manipulation check questions of frequency of purchase (α = .86; M = 2.48; SD = 2.17). Participants in the infrequently purchased items conditions were less familiar with the items they had to remember, consumed these products less frequently and bought these products less frequently (M = 1.47, SD = 2.07) than those in the frequently purchased items conditions (M = 3.47, SD = 1.79; F(1, 405) = 109.02, p b .01). Regarding the search strategy used, participants in the stimulus-based search strategy conditions reported that they browsed the categories on the website more (M = 4.12, SD = 1.51) than those in the memory-based shopping strategy conditions (M = 1.22, SD = 0.55; F(1, 404) = 434.16, p b .01) and in the control conditions (M = 1.78, SD = 1.21; F(1, 404) = 273.78, p b .01). Participants in the memory-based shopping strategy conditions reported that they browsed the categories less than those in the control conditions (F(1, 404) = 16.64, p b .01). Hypothesis tests As in Study 1a, we analyzed the data using a repeated measures ANOVA where the frequency of item purchase (frequently vs. infrequently purchased) and shopping strategy (memory-based vs. stimulus-based vs. control) were entered as between-participants factors. The number of items participants predicted they would remember (memory predictions) versus the number of items they correctly remembered (memory performance) was entered as a within-participant factor. We found only a three-way interaction between frequency of item purchase, shopping strategy, and memory predictions vs. performance (F(2, 401) = 2.79, p = .06, Fig. 5). We decomposed this three-way interaction into two-way interactions between frequency of item purchase and shopping strategy for each of the two dependent measures (predicted and actual memory performance) separately. For memory performance, we found no main effects of shopping strategy (p = .64) or of frequency of item purchase (p = .77), but, as expected, we found a significant interaction between these two factors (F(2, 401) = 3.49, p = .03). Participants remembered frequently purchased items better than infrequently purchased items in the memory-based search strategy condition (Mfreq = 7.48, SD = 2.54 vs. Minfreq = 6.61, SD = 2.70; F(1, 401) = 3.85, p = .05) and a similar amount of infrequently

and frequently purchased items in the control condition (Mfreq = 7.00, SD = 2.65 vs. Minfreq = 6.82, SD = 3.01; p = .69). Infrequently purchased items were marginally better remembered than frequently purchased items in the stimulus-based search strategy condition (Mfreq = 6.81, SD = 2.94 vs. Minfreq = 7.63, SD =1.88; F(1, 401) = 3.12, p = .08). For memory predictions, we only found that memory was predicted to be higher for frequently purchased items than for infrequently purchased items (Mfreq = 7.24, SD = 2.09 vs. Minfreq = 6.81, SD = 2.03; F(1, 401) = 4.27, p = .04). There was no main effect of shopping strategy (p = .69), and no interaction between these two factors (p = .59). Thus, as hypothesized, memory performance was lower for infrequently purchased items than for frequently purchased items in the memory-based search condition. However, in the stimulus-based search condition, infrequently purchased items were better remembered than frequently purchased items. We again find that the interaction effect of frequency of item purchase and search strategy on memory performance is not anticipated spontaneously in memory predictions. In both studies 1a and 1b, we find that the effect of frequency of item purchase is stronger in the memory-based condition (freq N infreq) than in the stimulus-based condition (freq b infreq). One potential explanation is that, according to dual-processes theories of recognition, “recognition memory must include a recollection process” (Dana, Reder, Arndt, & Park, 2006, p. 18). That is, knowing exactly the item one is searching for helps recognition memory. Thus, when participants are browsing the website in search for an item, having the item in memory helps the search as it may be the case of highly familiar words. Consistent with this, recent evidence shows that the recognition advantage of unfamiliar words is only significant when unfamiliar words are compared with words of average familiarity but not when unfamiliar words are compared with highly familiar words (Hemmer & Criss, 2013). This may explain why the effect of frequency of item purchase is stronger in the memory-based condition than in the stimulus-based condition. In the next study, the hypothesized underlying process for the interaction between the frequency of item purchase and search strategy (i.e., memory-based search's reliance on recall and stimulus-based search's reliance on recognition) is tested. We do this by using standard recall and recognition tasks.

Study 2: recall versus recognition Our predictions in studies 1a and 1b were based on the notion that the shopping strategies involving memory-based search versus stimulus-based search rely on recall versus recognition memory respectively. If this is indeed the case, we should find similar effects to the results revealed in studies 1a and 1b when we replace the manipulations of shopping strategy with standard measures of memory versus recognition. Thus, while in study 1 we instructed participants to either use the search function or browse the categories of the store to find the items, in study 2 we instructed participants to either recall the

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Number of items

*

9. 8. 7.

371

* 7.63

7.48

7.00 6.82

6.61

6.81

7.25

7.27

7.19 6.98

6.55

6.91

Frequently purchased items

6. 5. 4. 3.

Infrequently purchased items

2. 1. . Memory-based

Control

Stimulus-based Memory-based

Memory Performance

Control

Stimulus-based

Memory Predictions

Fig. 5. Study 1b results. Memory predictions and performance as a function of type of items and of search strategy. Note: *significant pairwise tests reported in the text.

items from their memory or identify the items on a list by clicking their name. Method Participants and design Eighty-three students at a major Southwest European university participated in the study in exchange for course credit (Mage = 22.49, SD = 0.88; 51% female). The key independent variables were frequency of item purchase (frequent vs. infrequent) and type of memory task (recall vs. recognition), both variables were manipulated between participants. We again measured both predicted and actual memory performance for each participant. Materials and procedure A pre-test was conducted to select frequently and infrequently purchased items. We initially included 20 grocery products we expected to be purchased frequently and 20 products we expected to be purchased infrequently in a pre-test. Those products were shown in random order to 11 respondents who were asked to indicate whether they use or buy the grocery product on a daily basis or only occasionally. Ten frequently purchased and ten infrequently purchased products with the highest percentage of frequent and of infrequent users were selected. The frequently purchased products and their percentage of identification as frequently purchased were: apples (100%), butter (100%), coke (82%), cookies (82%), ham (100%), lettuce (91%), olive oil (100%), shampoo (100%), spaghetti (100%), and tuna (91%). The infrequently purchased products and their percentage of identification as infrequently purchased were: coal (82%), mussels (91%), figs (91%), frozen cake (91%), muffin (91%), peach compote (100%), pistachios (91%), quince jelly (100%), red tea (100%), and sparkling water (91%). The convergence between respondents' identification of the selected items and the initial identification of those items was very strong (Cohen's K = 0.87). Kappa values higher than 0.81 indicate almost perfect agreement (Landis & Koch, 1977).

Participants were presented with either the frequently or the infrequently purchased items listed above and instructed to memorize them in order to remember them 5 min later. After memorizing the items, participants in the recognition condition answered the question “how many of the 10 products shown on the previous screen do you think you will be able to identify on a list of 100 items 5 minutes from now?” Participants in the recall condition answered the question “how many of the 10 products shown on the previous screen do you think you will be able to type out 5 minutes from now?” After participants indicated the number of products they thought they would remember, a time interval followed in which they completed a series of unrelated tasks. On average, participants took 3 min and 31 s to complete the unrelated tasks (SD = 1.79). Finally, participants in the recall condition were asked to type the products they could remember in a textbox, while participants in the recognition condition were asked to identify the products from a list of 100 products. Results and discussion We analyzed the data using a repeated measures ANOVA. Frequency of item purchase (frequent vs. infrequent) and type of memory task (recall vs. recognition) were entered as between-participants factors. The number of items participants predicted they would remember (memory predictions) versus the number of items participants correctly remembered (actual memory performance) was entered as a within-participant factor. We found a three-way interaction between frequency of item purchase, type of task, and memory predictions vs. performance (F(1, 79) = 13.53, p b .01, Fig. 6). We decomposed this three-way interaction into two-way interactions for memory predictions and memory performance. For memory performance, we found significant effects of type of task (Mrecall = 6.26, SD = 2.25 vs. Mrecognition = 8.69, SD =1.73; F(1, 79) = 32.81, p b .01), frequency of item purchase (Mfreq = 7.95, SD = 2.28 vs. Minfreq = 7.20, SD = 2.31; F(1, 79) = 3.61, p = .06), and, as expected, the

372

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Number 10. of items 9. 8.

*

*

8.65 8.73 7.65

7.11 6.47 6.47

7. 6.

5.42

5.95

Frequently purchased items

5.

Infrequently purchased items

4. 3. 2. 1. .

Recall

Recognition

Memory Performance

Recall

Recognition

take into account those effects when predicting their memory. We were also interested in participants use of shopping lists, as using a shopping list would be a practical and effective way to prevent forgetting to buy grocery items, therefore mitigating the effects of shopping strategy and item purchase frequency shown in the previous studies. We predicted that the more items participants thought they would forget the more likely they would indicate that they would use a shopping list. The test is important because we have assumed that memory predictions drive behavior.

Memory Predictions

Method Fig. 6. Study 2 results. Memory predictions and performance as a function of type of items and of tasks. Note: *significant pairwise tests reported in the text.

interaction between these two factors (F(1, 79) = 4.31, p =.04). We conducted separate contrasts to compare memory performance for frequently versus infrequently purchased items in recognition and in recall tasks. In the recall task, participants remembered more items in the frequently purchased item condition than in the infrequently purchased item condition (Mfreq = 7.11, SD = 2.33 vs. Minfreq = 5.42, SD = 1.86; F(1, 79) = 7.29, p b .01). In the recognition task, participants remembered a similar number of frequently and infrequently purchased items (Mfreq = 8.65, SD = 2.03 vs. Minfreq = 8.73, SD = 1.39; p = .90). The pattern was not anticipated in the participant's memory predictions. For memory predictions, we found that memory was predicted to be marginally higher for frequently purchased items than for infrequently purchased items (Mfreq = 7.12, SD = 2.22 vs. Minfreq = 6.20, SD = 2.08; F(1, 79) = 3.28, p = .07), no effect of the type of task (p = .48), and a marginally significant interaction between these two factors (F(1, 79) = 3.28, p = .07). In the recall task conditions, participants predicted that they would remember the same number of items in the frequently purchased items condition and the infrequently purchased items conditions (Mfreq = 6.47, SD = 2.14 vs. Minfreq = 6.47, SD = 1.95; p = 1). And, surprisingly, in the recognition task conditions, they predicted that they would remember more items in the frequently purchased items condition compared to the infrequently purchased item condition (Mfreq = 7.65, SD = 2.19 vs. Minfreq = 5.95, SD = 2.19; F(1, 79) = 7.16, p = .01). In sum, actual memory performance using standard recall and recognition tasks in study 2 closely mimicked the results for shopping strategies in studies 1a and 1 b, consistent with our proposition that the interaction between shopping strategy and item purchase frequency is driven by shopping strategies' reliance on recall versus recognition processes. As in studies 1a and 1b, the pattern of results for actual memory was not anticipated in participants' predictions of their later memory. Study 3a: naivety check Study 3a examined whether people are naïve about when they might forget. Our goal was to test whether people really do not know how shopping strategy and frequency of item purchase impact memory performance versus simply fail to spontaneously

Participants and design One-hundred and four individuals from a large metropolitan area in Southwest Europe reported their beliefs about their memory for grocery items. Relatives or acquaintances of our research assistant were invited over email to answer the survey online. Respondents were asked to distribute the link for the study to other shoppers. When the desired sample size was reached (about 100 responses, Mage = 39.12, SD = 13.68; 75% female), data collection stopped. Respondents reported that, on average, they shop for groceries 1.98 times per week, spend 35 min at the store per visit, and buy 21.31 products per visit of which 5.02 are unfamiliar. We defined unfamiliar items for participants as infrequently purchased and/or consumed items. In terms of shopping list use, 15.39% of respondents always use a shopping list and 13.46% reported that they never use a list. Thus, there is not only strong individual variation on the decision to use a shopping list, but also situational variation across shopping trips as most consumers don't have a rule regarding the use of a list. This is important because it highlights the fact that consumers make a decision about writing a list. It is not a habit to always or never use a list. Most respondents (63.46%) said that they sometimes forget to buy something they had intended to buy, some said that they often forget to buy something (29.81%), very few said that they always forget to buy something (3.85%), and only 2.88% said that they never forget. Thus, forgetting to buy is not uncommon for most of our respondents. Procedure Participants answered the questions online at their leisure. After completing the questions reported above they answered four questions regarding their shopping behavior. Specifically, we asked them the following: 1. when are you more likely to remember familiar items? (a. when searching for the product in most aisles in the supermarket; b. when trying to remember the product right before entering the supermarket; c. a mix of both aforementioned situations). Option “a” describes a stimulus-based search strategy. Option “b” describes a memory-based search strategy; 2. when are you more likely to remember unfamiliar items? (the same options as in the question above); 3. imagine you had to buy 10 items in the grocery store, how many items would you remember to buy?;

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

4. in the same visit to the grocery store, what is the likelihood you would use a shopping list? (a. high; b. low; c. not sure). Results First, we examined the answers to the items about the search strategy to remember frequently and infrequently purchased items. We used a multinomial logistic regression to test the effect of item purchase frequency (frequently vs. infrequently purchased) on the choice of shopping strategy (stimulus-based search vs. memory-based search vs. a mix of both strategies). We found a significantly different distribution for the choice of search strategy between frequently and infrequently purchased items (Wald χ2 (2) = 36.52, p = .01, see Fig. 7). Repeated measures multinomial logistic regressions using the mixed strategy (option c) as the reference category confirmed that respondents were more likely to select the stimulus-based search strategy as the best strategy to remember infrequently purchased items than to remember frequently purchased items (Wald χ2 (1) = 32.44, p b .01). Conversely, respondents were as likely to select the memory-based strategy as the best strategy to remember infrequently purchased items and frequently purchased items (Wald χ2 (1) = 0.12, p = .73). For infrequently purchased items, 58.65% of respondents selected the stimulus-based strategy, 15.38% selected the memory-based search strategy and 25.96% selected the mix of both strategies. For frequently purchased items, 16.34% selected the stimulus-based strategy, 39.42% selected the memory-based search strategy and 44.23% selected the mix of both strategies. The responses indicate that grocery shoppers, when explicitly asked, can anticipate the interaction effect of frequency of item purchase and search strategy on their memory performance. Regarding the self-perceived likelihood of using a shopping list in the hypothetical shopping task, 23.08% of respondents reported that it was high, 56.73% reported that they were not sure about it and 20.19% reported that it was low. We examined whether memory predictions in the hypothetical shopping task influence the decision to use a shopping list (− 1 = low likelihood, 0 = not sure, 1 = high likelihood). This is relevant because an effect of memory predictions on shopping list usage, which has been shown to reduce forgetting to buy needed items (Block & Morwitz, 1999), would evidence the Percentage of 100 respondents 90

16.3%

80 58.7%

70 60

stimulus-based strategy

44.2%

mix of both

50 40 30 20

26% 39.4%

10

memory-based strategy

15.4%

0

Frequently purchased items

Infrequently purchased items

Fig. 7. Study 3a results. Percentage of respondents that select each search strategy to remember frequently and infrequently purchased items.

373

important behavioral consequences of memory predictions. Helping consumers to more accurately predict their future memory performance might prevent some of the forgetting demonstrated in studies 1a, 1b, and 2, because consumers most likely to forget, particularly those using a shopping strategy based on memory-based search to remember infrequently purchased items, would be more likely to use a shopping list. A regression revealed an effect of memory predictions on the likelihood of using of a shopping list (B = − .17, SE = .03, p b .01). Specifically, the higher the number of items respondents thought they would remember, the less likely they thought they would use a shopping list. Study 3a shows that respondents who estimate their in-store memory performance to be less accurate indicate that they would be more likely to use a shopping list. Study 3a also shows that people can anticipate the interaction effect of frequency of item purchase and search strategy on their memory performance when they are explicitly asked about the direction of the effects. The result is in sharp contrast with the findings of studies 1a, 1b, and 2, in which participants did not anticipate the interaction effect between search strategy and frequency of item purchase. Our explanation is that metacognitive knowledge is not spontaneously activated when constructing a forecast. However, it is theoretically possible that these seemingly contradicting findings occurred because of the samples we used. The respondents in Study 3a were experienced grocery shoppers. In order to check whether the same results hold when the sample is kept constant, in study 3b we replicated study 3a with undergraduates. Study 3b: naivety check II Study 3b uses the same procedure and design as in Study 3a but with undergraduate students at a major university in Southwest Europe to test whether the results hold with a sample presumably less experienced with grocery shopping. Participants were 28 undergraduate students. (Mage = 23.43, SD = 1.26; 43% female). They reported that, on average, they shop for groceries 1.83 times per week, spend 35 min per visit, and buy 12.07 products per visit. In terms of shopping list use, 14.28% of respondents always use a shopping list and 21.42% never use a list. Most respondents said that they sometimes forget to buy something (53.57%), some said that they often forget to buy something (42.85%), very few said that they never forget (3.57%) and no participants said that they always forget. Thus, forgetting to buy is also a problem for our student population. We replicated Study 3a. We again found a significantly different distribution for the choice of search strategy between frequently and infrequently purchased items (Wald χ2 (2) = 9.70, p = .01, see Fig. 8). Repeated measures multinomial logistic regressions using the mixed strategy as the reference category confirmed that respondents were more likely to select the stimulus-based search strategy as the best strategy to remember infrequently purchased items than to remember frequently purchased items (Wald χ2 (1) = 6.98, p = .01). Conversely, respondents were as likely to select the memory-based strategy as the best strategy to remember infrequently purchased items

374 Percentage of respondents

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380 100

7%

90 80 43%

70

50%

60

stimulus-based strategy mix of both

50 40

25%

30

50%

20

memory-based strategy

25%

10 0

Frequently purchased items

Infrequently purchased items

Fig. 8. Study 3b results. Percentage of respondents that select each search strategy to remember frequently and infrequently purchased items.

and frequently purchased items (Wald χ2 (1) = 0.11, p = .74). For infrequently purchased items, 50% of respondents selected the stimulus-based strategy, 25% selected the memory-based search strategy and 25% selected the mix of both strategies. For frequently purchased items, 7.14% selected the stimulus-based strategy, 50% selected the memory-based search strategy and 42.86% selected the mix of both strategies. This corroborates our previous finding that when shopping strategy and frequency of item purchase are explicitly highlighted, people know that the stimulus-based search strategy is especially useful to remember infrequently purchased items. Regarding the use of a shopping list in the hypothetical shopping task, 42.86% of respondents reported that the likelihood of using a shopping list was high, 39.28% reported that it was medium and 17.86% reported that it was low. As was the case in study 3a, there is a negative effect of memory predictions on the likelihood of using a list (B = − .26, SE = .09, p = .01). The higher the number of items participants thought they would remember, the lower their reported likelihood of using a shopping list. One concern is the correlational nature of the effect of memory predictions on the use of a shopping list. To rule out this concern, we conducted an additional study where we manipulated the difficulty of the memory task (and hence memory predictions). This study was run in the laboratory and involved a behavioral measure of shopping list usage. We replicate the finding that when memory predictions are low people are more likely to use a shopping list. Thus, when asked explicitly, both undergraduates and experienced shoppers were aware that they are more likely to remember infrequently purchased items when using the stimulus-based search strategy. In addition, memory predictions influence the decision to use a shopping list. General discussion Summary and contributions Memory depends on the interaction between the type of items one intends to buy and the shopping strategy (relying on memory-based search vs. on stimulus-based search). Consumers are more likely to forget infrequently purchased items when

using the memory-based search strategy (Studies 1a and 1b) or performing recall tasks (Study 2). This effect is attenuated and can even be reversed when using the stimulus-based search strategy (Studies 1a and 1b) or performing recognition tasks (Study 2). If predictions about one's own memory were perfectly calibrated, participants could make informed decisions about whether or not to take measures to prevent forgetting, that is, whether to undo the interaction effect we documented. For instance, if memory predictions were calibrated, participants could rehearse items a bit more when they thought they might forget. In real life, shoppers could write a shopping list when they predict forgetting. Indeed, we empirically explored and found that memory predictions affect one such preventative measure, writing a shopping list (Studies 3a and 3b). However, we find that predictions are far from perfectly calibrated. Consumers do not spontaneously anticipate the interaction between frequency of item purchase and search strategy on memory performance (Studies 1a, 1b and 2). Only when we explicitly ask them about when they are more likely to forget frequently and infrequently purchased items do we find that participants are aware of what affects forgetting (Studies 3a and 3b). The present paper explains when and why consumers forget to buy the items they need. Previous research has shown that memory is a crucial factor for consideration and choice (Hauser & Wernerfelt, 1990; Nedungadi, 1990) without taking into account the interaction between item characteristics and consumer in-store behavior. Using an online shopping paradigm, we find that in-store memory depends on the interaction between type of item to be purchased and shopping strategy. Infrequently purchased items are more likely to be forgotten than frequently purchased items when using memory-based search (a recall-based strategy), but not when using stimulus-based search (a recognition-based strategy). Consumers, however, fail to anticipate this effect despite knowing the situations in which they are more likely to forget when they are explicitly asked about these variables. Identifying mistakes in predicting memory performance is a first step towards remedying the problem of forgetting. When memory predictions are wrong, consumers don't adequately prepare themselves for the shopping task. This paper substantiates previous calls to examine when and why consumers forget to buy (Bettman, 1979; Block & Morwitz, 1999; Krishnan & Shapiro, 1999). Bettman (1979) pointed as the first important area for future research the study of “factors differentially affecting recognition and recall” (p. 44). Bettman (1979) concludes his paper by stating that “research on when consumers use recognition or recall seems very important, since the properties of recognition and recall... differ” (p. 51). Since then, research has examined the implications of using recall versus recognition memory processes for the choices consumers make (i.e., memory-based vs. stimulus-based choices; Lynch & Srull, 1982; Rottenstreich et al., 2007; Sanbonmatsu & Fazio, 1990). This paper extends this examination to whether consumers remember to buy a product, after an intention to purchase has been set. We introduce two shopping strategies that rely on recall versus recognition memory processes (i.e., memory-based vs. stimulus-based search). Our studies show that the use of these

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

shopping strategies has consequences for what types of items are more likely to be forgotten. In an online shopping paradigm, participants are more likely to forget infrequently purchased products than frequently purchased ones when they use the search function to find the products they need by directly typing the product's name in the search bar (memory-based search), but not when they browse the categories of the online store in search for the products they need (stimulus-based search). This research also helps reconcile previous findings in the literature. Consumers are often more likely to consider frequently purchased brands than infrequently purchased ones (Hauser & Wernerfelt, 1990; Janiszewski et al., 2013; Nedungadi, 1990; Pechmann & Stewart, 1990). However, in some cases, consumers are at least as likely to consider infrequently purchased brands, such as those from low-market share brands or lower price alternatives, as frequently purchased ones: when items are displayed at end-of-aisles (Bemmaor & Mouchoux, 1991), when items have a high number of shelf facings (Chandon et al., 2009), when options are ranked by quality or price (Diehl et al., 2003), and when options are presented one at a time instead of all at once (Mogilner et al., 2013). Our findings may help reconcile these results. Infrequently purchased items benefit more from in-store exposure and thus are more likely to be considered when search costs are reduced. Our results also have implications for the accuracy of memory predictions. Previous work has shown that memory predictions are less accurate immediately after encoding than after a short delay (Rhodes & Tauber, 2011). This is because people are unable to anticipate forgetting something they have in mind. People think that what comes to mind easily at the time of memory predictions will also come to mind at the time of test (Kornell et al., 2011), thus, harboring a “stability bias” in memory (Kornell & Bjork, 2009). We add to this work by showing that memory predictions are also insensitive to the interaction between type of item and retrieval strategy (which in this context is the shopping strategy). In addition, we show that this occurs despite people knowing that those factors influence memory. People are aware of when they are most likely to forget, but still fail to spontaneously incorporate this knowledge into their forecasts and subsequent shopping list making decisions. This lack of spontaneous insight suggests that consumers do not appropriately adjust their shopping strategy for the type of items they intend to buy. However, the studies reported above do not explicitly test this assumption. Therefore, we conducted an additional study with 122 participants in which we find that the type of items that participants need to buy does not influence their search behavior. The procedure was very similar to study 1b except that we measured, instead of manipulated, search behavior. We used the number of categories that participants entered inside the store as a proxy for their search behavior. The more categories they entered, the more they used the stimulus-based search. We found no effect of the type of items (frequently vs. infrequently purchased) on memory predictions or on search behavior (ps N .41). In addition, we found an interaction between type of items and search behavior on memory

375

performance (F(1, 118) = 8.47, p b .01). The more categories participants entered inside the store, the lower was the difference in memory between frequently and infrequently purchased items. These exploratory results indicate that consumers do not spontaneously adjust their shopping strategy based on the type of items they intend to buy and yet the shopping strategy used interacts with the type of items to predict their memory performance.

Future research Future research may examine individual difference measures that predict the use of shopping strategies. For instance, recent research suggests that female consumers are more likely than males to rely on stimulus-based search (Heisz, Pottruff, & Shore, 2013). Thus, forgetting infrequently purchased products may be a more prevalent problem among men. Future research, may also examine individual difference factors that influence the effectiveness of each search strategy. For instance, recall performance declines more than recognition memory performance as consumers age (Danckaert & Craik, 2013). Thus, older consumers may benefit less from using a memory-based search strategy to buy frequently purchased items and perhaps should pay special attention not to forget this type of item. Furthermore, future research may investigate how to ensure that memory predictions are better calibrated. Other miscalibrations of future performance such as the planning fallacy (the belief that it takes less time to execute a task than it actually takes; Buehler, Griffin, & Peetz, 2010) and the illusion of explanatory depth (the belief that one knows more about how everyday objects work than what one actually knows; Fernbach, Sloman, Louis, & Shube, 2013), are reduced when participants are induced to process information more diligently to predict their future performance. For example, when asked to provide a mechanistic explanation, people realize that they do not know much about how everyday objects work (Alter, Oppenheimer, & Zemla, 2010). Thus, shifting information processing from an intuitive mode to a more reflective mode may help shoppers correctly estimate their future memory performance. Another open question is why consumers do not update their predictions on the next shopping trip. The issue of how reduced performance without a list impacts future decisions to use a list (i.e., the extent to which consumers learn from their mistake) should be examined in future research. We speculate that consumers learn insufficiently as they often misremember their forecasts as being consistent with their experience and thus fail to perceive the extent of their forecasting error (Meyvis, Ratner, & Levav, 2010). Future research may also examine other consequences than memory performance of using memory-based and stimulus-based search to shop for frequently and infrequently purchased items. Shoppers may find easier to shop for frequently purchased items using memory-based search and for infrequently purchased items using stimulus-based search. Therefore, the shopping experience is probably more efficient and perhaps also more pleasurable when shopping for frequently purchased items using memory-based

376

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

search and for infrequently purchased items using stimulus-based search. Managerial implications A limitation of our work is that we only tested our hypotheses in an online shopping context. Thus, future research may test our hypothesis in brick and mortar setting. In a brick and mortar shopping context, a shopping strategy relying on memory-based search would consist of consumers entering the store, trying to recall the items they need to buy, and moving directly to where they think the recalled items are located to place those items in their shopping carts. A shopping strategy relying on stimulus-based search would consist of consumers “walking the aisles” and counting on recognition memory to identify the needed items while moving by those items. To the extent that they would replicate in physical stores, our results offer suggestions to managers trying to assist consumers in remembering all of the items on their mental shopping lists. Recent research on shopping paths shows that shoppers often don't walk all the way down the aisles in the grocery store (Hui et al., 2009). Relatively few shoppers visit the middle of the aisle. Our results would suggest that supermarkets place the items that shoppers frequently buy in the middle of the aisle and the items they infrequently buy at the ends. As a result, when shoppers go to ‘pick’ an item they frequently buy they are more likely to be exposed to other products including the ones they infrequently buy and therefore to remember that they need to buy these products. Other strategies that motivate shoppers to walk the aisles of the store may also be effective at helping their memory for infrequently purchased items. For instance, placing a promotion in the middle of the aisle may motivate shoppers to move to the promoted item and be exposed to other products on the way. Managers may also help shoppers' memory by encouraging the use of shopping lists. The Coles online store for instance allows shoppers to prepare a shopping list for later purchase (see Fig. 1). Grocery retailers can also help shoppers' memory by cuing them to use memory-based search to buy frequently purchased items and stimulus-based search to buy infrequently purchased items. For instance, Safeway's online store allows shoppers to “Shop By History” and to “Shop By Aisle”. Safeway could instruct shoppers to find the products they frequently buy using the “Shop By History” option and the products they infrequently buy using the “Shop By Aisle” option. One problem is that both options are stimulus-based. In the “Shop By History” option, previous purchases are presented to shoppers and in the “Shop By Aisle” option all items of the store are sorted in categories. It may be very time-consuming for consumers to go through all items of an online grocery store or even all the items they ever bought to find the products they need. Safeway does a good job in the “Shop By Aisle” option by facilitating stimulus-based search without making the shopper go through all the items: a smart category structure that allows consumers to easily skip whole categories (i.e. the pets category), parts of categories (sodas, light beer), and that clusters some items in specific subcategories (seasonal fruit,

lactose free milk). Perhaps Safeway should make the “Shop By History” option more memory-based by for instance asking shoppers to list the products they need before presenting them with the items they previously bought. Conclusion Although our empirical strategy centered on shopping for groceries, implications of our findings should extend to other tasks in which memory is crucial. For example, consumers need to remember to pay the bills, book a flight, save money, and take medication, among other tasks. Our findings suggest that consumers are more likely to remember frequently performed tasks from their memory and infrequently performed tasks with the help of stimulus cues. We also show that consumers fail to anticipate that they would forget. When people are not accurate about how well they will remember, they cannot adequately prepare memory aids. Thus, for instance, when planning to go shop for groceries, consumers might be better off erring on the side of caution and using a simple rule: Don't ask yourself if you'll remember, just write a shopping list. Appendix A. Additional information about our studies Study 1a Phase 1: Presentation of items to participants and memory predictions. Verbatim instructions in the memory-based condition (A), in the control condition (B) and in the stimulus-based condition (C). Either infrequently or frequently purchased items were presented. Infrequently purchased items are in italic. A. In the next experiment, we are investigating whether people are able to remember to buy the things they need to buy. 5 min from now, we'll ask you to enter an online grocery store and to buy the 10 fruits & vegetables below by typing their names into the store's search engine. Therefore, please memorize these products. 1. Apple; 2. Banana; 3. Broccoli; 4. Lettuce; 5. Papaya; 6. Potato; 7. Pumpkin; 8. Strawberry; 9. Tomato; 10. Watermelon. 1. Avocado; 2. Beetroot; 3. Celery; 4. Coconut; 5. Fig; 6. Mushroom; 7. Passion fruit; 8. Pineapple; 9. Plums; 10. Rhubarb. How many of the 10 products above do you think you will remember and be able to type on the search engine of the online grocery store 5 min from now? B. In the next experiment, we are investigating whether people are able to remember to buy the things they need to buy. 5 min from now, we'll ask you to enter an online grocery store and to buy the 10 fruits & vegetables below. Therefore, please memorize these products. 1. Apple; 2. Banana; 3. Broccoli; 4. Lettuce; 5. Papaya; 6. Potato; 7. Pumpkin; 8. Strawberry; 9. Tomato; 10. Watermelon.

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

1. Avocado; 2. Beetroot; 3. Celery; 4. Coconut; 5. Fig; 6. Mushroom; 7. Passion fruit; 8. Pineapple; 9. Plums; 10. Rhubarb. How many of the 10 products above do you think you will remember and be able to buy in the online grocery store 5 min from now? C. In the next experiment, we are investigating whether people are able to remember to buy the things they need to buy. 5 min from now, we'll ask you to enter an online grocery store and to buy the 10 fruits & vegetables below by browsing the categories in the online store. Therefore, please memorize these products. 1. Apple; 2. Banana; 3. Broccoli; 4. Lettuce; 5. Papaya; 6. Potato; 7. Pumpkin; 8. Strawberry; 9. Tomato; 10. Watermelon. 1. Avocado; 2. Beetroot; 3. Celery; 4. Coconut; 5. Fig; 6. Mushroom; 7. Passion fruit; 8. Pineapple; 9. Plums; 10. Rhubarb. How many of the 10 products above do you think you will remember and be able to identify on the categories of the online grocery store 5 min from now? Phase 2: Time delay of 8 min and 9 s on average. Phase 3: Shopping task instructions. Verbatim instructions in the memory-based condition (A), in the control condition (B) and in the stimulus-based condition (C). A. Now, we ask you to find the products you were asked to remember. Please enter the Coles online store and search for the products by typing the product names in the search bar of the website. Please do not browse the website to find the products. It's important that you only use the search bar of the website to find the products by typing their names. The website of the store is the following: http://shop.coles.com.au/ online/national. As you find the products you need to buy, please type below their names and their codes. The codes are available once you click on the items B. Now, we ask you to find the products you were asked to remember. Please enter the Coles online store and search for the products. The website of the store is the following: http://shop.coles.com.au/online/national. As you find the products you need to buy, please type below their names and their codes. The codes are available once you click on the items. C. Now, we ask you to find the products you were asked to remember. Please enter the Coles online store and search for the products by looking for the products in the fruits & vegetables category of the website. Please do not use the search bar of the website to find the products. It's important that you only browse the website in the Shop by Category option to find the products by looking for their names in the fruits & vegetables category. The website of the store is the following: http://shop.coles.com.au/online/national. As you find the products you need to buy, please

377

type below their names and their codes. The codes are available once you click on the items. Phase 4: Manipulation check of frequency of item purchase: Please answer the following questions about the products you were supposed to remember. 1. Please indicate how familiar are you with the products you were supposed to remember: − 5 (very unfamiliar) − 4 − 3 − 2 − 1 0 1 2 3 4 5 (very familiar) 2. Please indicate how frequently you consume the products you were supposed to remember: − 5 (very infrequently) − 4 − 3 − 2 − 1 0 1 2 3 4 5 (very frequently) 3. Please indicate how frequently you buy the products you were supposed to remember: − 5 (very infrequently) − 4 − 3 − 2 − 1 0 1 2 3 4 5 (very frequently) Phase 5: Manipulation check of search strategy: Please indicate how you proceeded in finding the products in terms of whether you used the search function or browsed the fruit & vegetables category: 1. I only used the search function; 2. I mostly used the search function but also browsed the category a little; 3. I used the search function and browsed the category about the same; 4. I mostly browsed the category but also used the search function; 5. I only browsed the category. Study 1b Phase 1: Instruction manipulation check. Verbatim measure: Recent research on decision making shows that choices are affected by context. Differences in how people feel, their previous knowledge and experience, and their environment can affect choices. To help us understand how people make decisions, we are interested in information about you. Specifically, we are interested in whether you actually take the time to read the directions; if not, some results may not tell us very much about decision making in the real world. To show that you have read the instructions, please ignore the question below about how you are feeling and instead check only the “none of the above” option as your answer. Thank you very much. Please check all words that describe how you are currently feeling: 1. Interested; 2. Distressed; 3. Excited; 4. Upset; 5. Strong; 6. Guilty;… 21. None of the above. Phase 2: Presentation of items to participants and memory predictions. The same as in Study 1a except that frequently purchased item number 5 was orange instead of papaya.

378

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Phase 3: Time interval of 10 min. Phase 4: Shopping task instructions. The same as in Study 1a. Phase 5: Manipulation check of frequency of item purchase. The same as in Study 1a. Phase 6: Manipulation check of search strategy. The same as in Study 1a. Study 2 Phase 1: Presentation of items to participants. In the next experiment, we are investigating whether people are able to remember to buy the things they need to buy. Therefore, please memorize the 10 products below. After 5 min, at the end of the study session, you will have to remember the products presented below. 1. Lettuce; 2. Apples; 3. Olive Oil; 4. Butter; 5. Coke; 6. Ham; 7. Spaghetti; 8. Cookies; 9. Tuna; 10. Shampoo. 1. Red tea; 2. Mussels; 3. Figs; 4. Muffin; 5. Quince Jelly; 6. Peach Compote; 7. Pistachios; 8. Frozen Cake; 9. Sparkling Water; 10. Coal. Phase 2: Memory predictions. Verbatim measure in the recall condition (A) and in the recognition condition (B): A. How many of the 10 products shown on the previous screen you think you will remember and be able to type out 5 min from now? B. How many of the 10 products shown on the previous screen you think you will remember and be able to identify on a list of 100 products 5 min from now? Phase 3: Time delay of 3 min and 31 s on average. Phase 4: Memory task. Verbatim instructions in the recall condition (A) and in the recognition condition (B): A. Memory test. Please type out below the products you were asked to remember. B. Memory test. Please identify in the list below the products you were asked to remember.

Bananas Carrot Peach Apples Orange Mango Melon Broccolli Rucola Lettuce Tomato Spinach Pumpkin Mushroom

Food bags Toilet rolls Batteries Rice Spaghetti Granola Cornflakes Peanut butter Strawberry jam Quince jelly Cheesecake Apple pie Cookies Candies

Figs Olive oil Vinegar Dressing Ketchup Sushi Tomato soup Mustard Baguette Bagels Waffles Muffin Brown bread Brioche Croissant Ciabatta Ricotta Gouda Butter Milk Low-fat milk Sorbet Frozen pizza Frozen cake Ice cream Iced tea Beer Red wine Coke Sparkling water Energy drink Coffee Cappuccino Green tea Black tea Red tea

Peach compote Strawberry tart Varnish Toilet duck Liquid detergent Coal Blender Spatula Sausage Entrecote Beef Ham Salami Meatballs Salmon Tuna Mussels Chicken Turkey Pretzels Crisps Pistachios Peanuts Popcorn Crackers Lipstick Paracetamol Calcium Sunscreen Soap Shampoo Conditioner Tissues Deodorant Toothbrush Toothpaste

Study 3a and 3b Phase 1: Measurement of grocery shopping behavior. a. How many times per week you go to the supermarket? b. How much time on average you take each time you go to the supermarket? (in minutes) c. How many products on average you buy each time you go to the supermarket? d. What percentage of the products you buy are familiar products that you buy almost every week? (Only measured in Study 3a.) e. How often you use a shopping list? 1. I always use it; 2. I most often use it; 3. I sometimes use it; 4. I never use it. f. How often you forget to buy something you needed at the supermarket? 1. Never forget to buy any product; 2. Usually I don't forget to buy any product; 3. Most often I forget to buy a product; 4. I always forget to buy a product. Phase 2: Measurement of participants' theories about forgetting. For the next questions, please consider the following definitions: Familiar products: products you frequently consume. For example: apples, rice…

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Unfamiliar products: products you don't frequently consume. For example: figs, red tea… a. when are you more likely to remember familiar items? (a. when searching for the product in most aisles in the supermarket; b. when trying to remember the product right before entering the supermarket; c. a mix of both aforementioned situations). b. when are you more likely to remember unfamiliar items? (a. when searching for the product in most aisles in the supermarket; b. when trying to remember the product right before entering the supermarket; c. a mix of both aforementioned situations); Phase 3: Hypothetical shopping experience a. imagine you had to buy 10 items in the grocery store, how many items would you remember to buy?; b. in the same visit to the grocery store, what is the likelihood you would use a shopping list? (in Study 3a: a. high; b. low; c. not sure; in Study 3b: a. high; b. medium; c. low). Appendix B. Results including participants excluded in Study 1b When all the 463 participants are included, the three-way interaction between frequency of item purchase, shopping strategy and memory predictions vs. performance was not significant (F(2, 457) = 1.26, p = .28). For memory performance, we found no main effects of shopping strategy (p = .70) or of frequency of item purchase (p = .25), and a marginally significant interaction between these two factors (F(2, 457) = 2.51, p = .08). Participants remembered frequently purchased items better than infrequently purchased items in the memory-based search strategy condition (Mfreq = 7.44, SD = 2.51 vs. Minfreq = 6.38, SD = 2.86; F(1, 457) = 6.01, p = .01), but not in the control condition (Mfreq = 6.91, SD = 2.79 vs. Minfreq = 6.80, SD = 2.92; p = .79), and in the stimulus-based search strategy condition (Mfreq = 6.96, SD = 2.90 vs. Minfreq = 7.25, SD = 2.32; p = .51). For memory predictions, we found a main effect of frequency of item purchase (Mfreq = 7.25, SD = 2.09 vs. Minfreq = 6.77, SD = 2.10; F(1, 457) = 5.89, p = .02), no main effect of shopping strategy (p = .50), and no interaction between these two factors (p = .51). References Alter, A. L., Oppenheimer, D. M., & Zemla, J. C. (2010). Missing the trees for the forest: A construal level account of the illusion of explanatory depth. Journal of Personality and Social Psychology, 99, 436–451. Bell, D., Corsten, D., & Knox, G. (2011). From point-of-purchase to path-topurchase: How preshopping factors drive unplanned buying. Journal of Marketing, 75, 31–45. Bemmaor, A. C., & Mouchoux, D. (1991). Measuring the short-term effect of in-store promotion and retail advertising on brand sales: A factorial experiment. Journal of Marketing Research, 28, 202–214.

379

Benjamin, A., Bjork, R. A., & Schwartz, B. L. (1998). The mismeasure of memory: When retrieval fluency is misleading as a metamnemonic index. Journal of Experimental Psychology: General, 127, 55–68. Bettman, J. R. (1979). Memory factors in consumer choice: A review. Journal of Marketing, 43, 37–53. Block, L., & Morwitz, V. G. (1999). Shopping lists as an external memory aid for grocery shopping: Influences on list writing and list fulfillment. Journal of Consumer Psychology, 8, 343–376. Buehler, R., Griffin, D., & Peetz, J. (2010). The planning fallacy: Cognitive, motivational, and social origins. Advances in Experimental Social Psychology, 43, 1–62. Chandon, P., Hutchinson, J. W., Bradlow, E. T., & Young, S. (2009). Does instore marketing work? Effects of the number and position of shelf facings on brand attention and evaluation at the point of purchase. Journal of Marketing, 73, 1–17. Dana, R. A., Reder, L. M., Arndt, J., & Park, H. (2006). Models of recognition: A review of arguments in favor of a dual-process account. Psychonomic Bulletin & Review, 13, 1–21. Danckaert, S. L., & Craik, F. I. M. (2013). Does aging affect recall more than recognition memory? Psychology and Aging, 28, 902–909. Diehl, K., Kornish, L. J., & Lynch, J. G., Jr. (2003). Smart agents: When lower search costs for quality information increase price sensitivity. Journal of Consumer Research, 30, 56–71. Dobbins, I. G., Kroll, N. E. A., Yonelinas, A. P., & Liu, Q. (1998). Distinctiveness in recognition and free recall: The role of recollection in the rejection of the familiar. Journal of Memory and Language, 38, 381–400. Dunlosky, J., & Bjork, R. A. (2008). The integrated nature of metamemory and memory. In J. Dunlosky, & R. A. Bjork (Eds.), Handbook of metamemory and memory (pp. 11–28). Mahwah, NJ: Erlbaum. Fernbach, P., Sloman, S., Louis, R., & Shube, J. (2013). Explanation fiends and foes: How mechanistic detail determines understanding and preference. Journal of Consumer Research, 39, 1115–1131. Food Marketing Institute (2012). U.S. grocery shopper trends. Accessed on: July 27th, 2014. Available at: http://www.icn-net.com/docs/12086_FMIN_ Trends2012_v5.pdf Gilbride, T. J., Inman, J. J., & Stilley, K. M. (2015). The role of within-trip dynamics in unplanned versus planned purchase behavior. Journal of Marketing, 79, 57–73. Granbois, D. H. (1968). Improving the study of customer in-store behavior. Journal of Marketing, 32, 28–33. Gregg, V. H. (1976). Word frequency, recognition, and recall. In J. Brown (Ed.), Recall and recognition. London: Wiley. Hauser, J. R., & Wernerfelt, B. (1990). An evaluation cost model of consideration sets. Journal of Consumer Research, 16, 393–408. Heisz, J. J., Pottruff, M. M., & Shore, D. I. (2013). Females scan more than males: A potential mechanism for sex differences in recognition memory. Psychological Science, 24, 1157–1163. Hemmer, P., & Criss, A. H. (2013). The shape of things to come: Evaluating word frequency as a continuous variable in recognition memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39, 1947–1952. Hui, S., Bradlow, E., & Fader, P. (2009). Testing behavioral hypotheses using an integrated model of grocery store shopping path and purchase behavior. Journal of Consumer Research, 36, 478–493. Hui, S., Huang, Y., Suher, J., & Inman, J. (2013). Deconstructing the ‘first moment of truth’: Understanding unplanned consideration and purchase conversion using in-store video tracking. Journal of Marketing Research, 50, 445–462. Hui, S., Inman, J., Huang, Y., & Suher, J. (2013). Estimating the effect of travel distance on unplanned spending: Applications to mobile promotion strategies. Journal of Marketing, 77, 1–16. Inman, J. J., Winer, R. S., & Ferraro, R. (2009). The interplay between category characteristics, customer characteristics, and customer activities on in-store decision making. Journal of Marketing, 73, 19–29. Iyer, E. S. (1989). Unplanned purchasing: Knowledge of shopping environment and time pressure. Journal of Retailing, 65, 40–57. Janiszewski, C., Kuo, A., & Tavassoli, N. T. (2013). The influence of selective attention and inattention to products on subsequent choice. Journal of Consumer Research, 39, 1258–1274.

380

D. Fernandes et al. / Journal of Consumer Psychology 26, 3 (2016) 363–380

Kintsch, W. (1967). Memory and decision aspects of recognition learning. Psychological Review, 74, 496–504. Koriat, A., Bjork, R. A., Sheffer, L., & Bar, S. (2004). Predicting one's own forgetting: The role of experience-based and theory-based processes. Journal of Experimental Psychology: General, 133, 643–656. Kornell, N., & Bjork, R. A. (2009). A stability bias in human memory: Overestimating remembering and underestimating learning. Journal of Experimental Psychology: General, 138, 449–468. Kornell, N., Rhodes, M. G., Castel, A. D., & Tauber, S. K. (2011). The ease-ofprocessing heuristic and the stability bias: Dissociating memory, memory beliefs, and memory judgments. Psychological Science, 22, 787–794. Krishnan, S., & Shapiro, S. (1999). Prospective and retrospective memory for intentions: A two-component approach. Journal of Consumer Psychology, 8, 141–166. Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159–174. Lynch, J. G., Jr., & Srull, T. K. (1982). Memory and attentional factors in consumer choice: Concepts and research methods. Journal of Consumer Research, 9, 108–128. Meyvis, T., Ratner, R., & Levav, J. (2010). Why don't we learn to accurately forecast feelings? How misremembering our predictions blinds us to past forecasting errors. Journal of Experimental Psychology: General, 139, 579–589. Mogilner, C., Shiv, B., & Iyengar, S. (2013). Eternal quest for the best: Sequential (vs. simultaneous) option presentation undermines choice commitment. Journal of Consumer Research, 39, 1300–1312. Nedungadi, P. (1990). Recall and consumer consideration sets: Influencing choice without altering brand evaluations. Journal of Consumer Research, 17, 263–276. New York Times (2011). Survey says consumers miss the mark with holiday manners and forget batteries. The New York Times.

Oppenheimer, D. M., Meyvis, T., & Davidenko, N. (2009). Instructional manipulation checks: Detecting satisficing to increase statistical power. Journal of Experimental Social Psychology, 45, 867–872. Pechmann, C., & Stewart, D. W. (1990). The effects of comparative advertising on attention, memory, and purchase intentions. Journal of Consumer Research, 17, 180–191. Reder, L. M., Nhouyvanisvong, A., Schunn, C. D., Ayers, M. S., Angstadt, P., & Hiraki, K. (2000). A mechanistic account of the mirror effect for word frequency: A computational model of RK judgments in a continuous recognition paradigm. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 294–320. Rhodes, M., & Tauber, S. (2011). The influence of delaying judgments of learning on metacognitive accuracy: A meta-analytic review. Psychological Bulletin, 137, 131–148. Rottenstreich, Y., Sood, S., & Brenner, L. (2007). Feeling and thinking in memory-based versus stimulus-based choices. Journal of Consumer Research, 33, 461–469. Sanbonmatsu, D. M., & Fazio, R. H. (1990). The role of attitudes in memorybased decision making. Journal of Personality and Social Psychology, 59, 614–622. Schwartz, B. L. (1994). Sources of information in metamemory: Judgments of learning and feeling of knowing. Psychonomic Bulletin and Review, 1, 357–375. Stilley, K. M., Inman, J. J., & Wakefield, K. L. (2010). Planning to make unplanned purchases? The role of in-store slack in budget deviation. Journal of Consumer Research, 37, 264–278. Thomas, A., & Garland, R. (2004). Grocery shopping: List and non-list usage. Marketing Intelligence & Planning, 22, 623–635.