Bottomless bowls why visual cues

This idea formed a core component of the well-known diet book Mindless Eating, and even won an Ig Nobel prize in Last week James Heathers, a postdoctoral researcher at Northeastern University in Boston, made public the results from a statistical reanalysis of data from the bottomless bowls study that call into question the veracity of the results. Heathers used a technique known as SPRITE short for Sample Parameter Reconstruction via Iterative Techniques to investigate whether the descriptive statistics reported in the paper the number of data points, means and standard deviations could possibly exist, given a assumptions about the distribution of the underlying data, and b restrictions on the data that are outlined in the paper.

A fuller explanation of SPRITE can be found here , but essentially, the idea is that you give the program a mean, standard deviation and some restrictions say, ten data points have to have a value of 3 , and it produces a set of possible histograms of data that could produce those values.

The hope is that some of these histograms follow sensible rules about data — are they normally distributed? Are the minimum and maximum values realistic? Unfortunately, if that cue is inaccurate, it could un- knowingly lead one to overeat. This suggests the following hypothesis: altering visual cues of how much is eaten will Figure 1: Mediated impact of portion size on consumption influence intake H1.

Visual cues can lead a person to underestimate how much they have consumed or to overestimate how much they have consumed. This can lead one to consume either more than they intended or less than they intended. For these people, there is a visual cue or amount of liquid they pour into short, wide drinking glasses benchmark they have established a clean plate , and they compared with tall, narrow glasses that hold the same vol- eat until they either reach that benchmark or until they are ume Similarly, a size-contrast illusion could lead a otherwise sated.

Just as with those people who instead intend to eat one-half or three-quarters of what they are person to underestimate and overconsume the amount of served, all are likely to overconsume if given a larger potion food on a large plate or to overestimate and underconsume size at the outset. Even if they are subsequently unable to the amount of food on a small plate. Such estimates, along with feelings of the point where they might have with a smaller, but still satiation, can be influenced by the volume of food they unconstrained, supply.

As with normative benchmarks in believed they consumed If people do not believe that other situations 15 , the influence of this visual cue may be they consumed a relatively larger volume of food, they may relatively automatic and may occur without deliberation.

While these biased visual cues lead people to unknowingly overeat, they Consumption Monitoring in Distracting Environments should do so without having a commensurate impact on Figure 1 illustrates that consumption norms and expecta- consumption estimates or on satiation. A second hypothesis tions may mediate or explain at least some part of the related to visual cues is the following: when given inaccu- impact portion size has on consumption. Closely tion estimation and perception of satiety will be more in- monitoring how much food one consumes reduces discrep- fluenced by the biased visual cue than by how much they ancies between perceived and actual consumption.

Unfor- actually consume H2. As a result, their estimate of how much they have eaten. As how much they actually ate. Figure 1 suggests, not being able to effectively monitor how much they consume can lead people to rely more heavily on easy-to-monitor visual cues that are related to their con- sumption expectations. The early work of Pudel and Oetting Research Methods and Procedures 16 showed this when participants who drank soup through This study examined whether altering a visual cue of how a tube drank less when they had visual contact with the soup much is eaten would influence intake and whether this than when they did not.

On the day ounce bowl of soup and asked to indicate what percent- and time they were assigned, they were met by a research age of the bowl they thought they would consume during a assistant in an adjoining room at the appointed time. To examine this in a bit more detail, soup used in the study were different colors either green or blue bowl sizes were adjusted in a cafeteria, and bowl waste was balanced across both conditions , and because they were measured.

In combination, this helped to suggest that two a meal-type setting. During each session, two participants common target levels may be one-half of the bowl and all of received the treatment self-refilling bowl, and two re- the bowl when consuming soup. A self-refilling bowl of ceived the control normal bowl. Participants were as- signed random seating. They were instructed to eat the soup soup eliminated the external cue otherwise provided by an and not to move the placement of the bowl from its predes- empty bowl.

Based on work by Schulundt et al. Kretsch et al. After this, they Participation was before starting the study, whether they generally try to eat voluntary.

The study had Institutional Review Board ap- until their bowl is empty, and whether they believed their proval, and participants were treated in accordance with consumption was influenced by the presence of others.

American Psychological Association guidelines. This ex- Consistent with Rolls 25 and Inman 26 , the satiety periment was a between-subject design with one factor and questions were asked using semantic differential scales two visibility levels: 1 accurate visual cue normal bowl e.

Although premeal hunger ratings are important, it was a concern that taking these Procedure measures would bias the results by cueing respondents to Individuals were recruited by a recruiter or in response to the issue of consumption volume. For this reason, respon- a flyer.

Individuals were informed that their participation dents were randomized to be certain that different degrees required eating a soup-only lunch and completing a ques- of hunger would be randomly and evenly distributed across tionnaire. A reminder e-mail was sent before the partici- the two conditions. To reduce the artificiality of the study and to introduce a The actual volume of soup consumed was determined by degree of distraction, subjects were run in groups of four.

It comparing the combined weight of the soup remaining in is important to acknowledge that social facilitation can the caldron, the tubing, and in the soup bowl with the soup influence consumption 22 , leading people to eat more with that was originally available.

The difference indicated the familiar companions 23,24 and to eat less with strangers. Because the objective of this study was to compare the two conditions self-refilling vs. This involved a heavy 1. Self-reported measures of social facilitation and famil- restaurant-style dining table 4.

Two soup bowls were between these conditions. Six quarts of soup were required for Recall that it was expected that those who had unknow- proper function of the apparatus. For the ratio of soup in the ingly been given the self-refilling soup bowls the biased pot to soup in the bowl to remain constant, the pot was visual cue would eat more than those given normal soup slightly elevated above the bowl.

This ensured that the soup bowls H1 , but they would not estimate themselves to flowed at a slow but constant rate. A manual drain was having eaten any more, nor would they perceive themselves connected to the conduit underneath the table for draining to be more sated H2. Both hypotheses were confirmed. This difference represented an bowls that held 18 ounces Sex and BMI were not significant when included as Under each of the two self-refilling bowls, 2-cm holes had covariates, indicating that this effect was consistent across been drilled through the 4.

Through this men and women and across people with different BMIs. For measures of consumption [5. However, when included to- bowls through a bayonet mounting. An the soup bowls through a gravity-feed mechanism. As a person ate from one of these bowls, the bowl would It was believed that participants used the fill level of soup slowly and almost imperceptibly refill itself.

The refill rate in the bowl as an indication of how much they would could be modified. If it refilled too slowly, people could consume. This tendency was observed in a pilot study and conceivably finish their soup before it refilled. To insure the context of this study, yet the reference point it presents that those in the control condition would have the opportu- can still influence consumption.

In this way, every person was able it. Those eating from normal bowls be- nificance. As expected, there was no difference between lieved they had eaten In contrast, conditions. That is, there were no differences between ret- those eating from self-refilling bowls believed they had rospective measures of consumption [5.

Obesity Research. ISSN Retrieved Associative learning and conditioning theory : human and non-human applications. Oxford University Press. ISBN OCLC The New York Times. Facebook Page. Category : Eating behaviors of humans. Tools What links here.