ArticlePDF Available

Losing a Dime With a Satisfied Mind: Positive Affect Predicts Less Search in Sequential Decision Making

Authors:

Abstract and Figures

We investigated the contribution of cognitive ability and affect to age differences in sequential decision making by asking younger and older adults to shop for items in a computerized sequential decision-making task. Older adults performed poorly compared to younger adults partly due to searching too few options. An analysis of the decision process with a formal model suggested that older adults set lower thresholds for accepting an option than younger participants. Further analyses suggested that positive affect, but not fluid abilities, was related to search in the sequential decision task. A second study that manipulated affect in younger adults supported the causal role of affect: Increased positive affect lowered the initial threshold for accepting an attractive option. In sum, our results suggest that positive affect is a key factor determining search in sequential decision making. Consequently, increased positive affect in older age may contribute to poorer sequential decisions by leading to insufficient search. (PsycINFO Database Record (c) 2012 APA, all rights reserved).
Content may be subject to copyright.
Psychology and Aging
Losing a Dime With a Satisfied Mind: Positive Affect
Predicts Less Search in Sequential Decision Making
Bettina von Helversen and Rui Mata
Online First Publication, March 26, 2012. doi: 10.1037/a0027845
CITATION
von Helversen, B., & Mata, R. (2012, March 26). Losing a Dime With a Satisfied Mind:
Positive Affect Predicts Less Search in Sequential Decision Making. Psychology and Aging.
Advance online publication. doi: 10.1037/a0027845
Losing a Dime With a Satisfied Mind:
Positive Affect Predicts Less Search in Sequential Decision Making
Bettina von Helversen and Rui Mata
University of Basel and Max Planck Institute for Human Development, Berlin, Germany
We investigated the contribution of cognitive ability and affect to age differences in sequential decision
making by asking younger and older adults to shop for items in a computerized sequential decision-
making task. Older adults performed poorly compared to younger adults partly due to searching too few
options. An analysis of the decision process with a formal model suggested that older adults set lower
thresholds for accepting an option than younger participants. Further analyses suggested that positive
affect, but not fluid abilities, was related to search in the sequential decision task. A second study that
manipulated affect in younger adults supported the causal role of affect: Increased positive affect lowered
the initial threshold for accepting an attractive option. In sum, our results suggest that positive affect is
a key factor determining search in sequential decision making. Consequently, increased positive affect in
older age may contribute to poorer sequential decisions by leading to insufficient search.
Keywords: decision making, sequential choice, aging, satisfaction, positive affect
Suddenly it happened, I lost every dime/But I’m richer by far with a
satisfied mind.
—J. H. Hayes/Jack Rhodes
From finding mates to apartments, many decisions people face
are sequential. Sequential decisions are often difficult because one
is forced to evaluate options on the spot and does not have the
luxury to change his mind: A partner will likely move on if you
court others, or a landlord will pick the next interested tenant if you
hesitate to take an apartment. The trick lies in neither stopping the
search too early nor too late so as not to miss out on the best
partner or apartment. One token of this class of problems is the
well-known secretary problem in which decision makers must
select the best job candidate out of a sequentially presented pool of
applicants without any prior knowledge about the distribution of
the applicants’ quality (Ferguson, 1989; Gilbert & Mosteller,
1966). The options are presented in a random order, and an option
that has been rejected cannot be recalled at a later time. The
optimal solution to this problem can be described by a simple
threshold strategy (Ferguson, 1989). According to the threshold
strategy, the decision maker searches through a number of options
to gain experience about the possible candidates’ quality. After
enough experience has been gained, a threshold is set equivalent to
the best option seen thus far and the next option that exceeds the
threshold is chosen. The threshold strategy describes well how
individuals solve the secretary problem, although people tend to
have lower thresholds and thus search less than the optimal strat-
egy would prescribe (Bearden, Rapoport, & Murphy, 2006; Seale
& Rapoport, 1997, 2000; von Helversen, Wilke, Johnson, Schmid,
& Klapp, 2011; but see Zwick, Rapoport, Lo, & Muthukrishnan,
2003). But do younger and older adults differ in their sequential
decision making?
Aging is associated with decline in cognitive abilities potentially
relevant for decision making. Older adults seem to perform worse
in a number of decision tasks because of cognitive limitations
(Bruine de Bruin, Parker, & Fischoff, 2010; Finucane & Gullion,
2010; Mata, von Helversen, Karlsson, & Cu¨pper, 2011; Mata, von
Helversen, & Rieskamp, 2010; Mata, Schooler, & Rieskamp,
2007). At the same time, aging is associated with affective and
motivational changes found to affect decision making by influenc-
ing predecisional information search (Löckenhoff & Carstensen,
2007), postchoice memory (Mather, Knight, & McCaffrey, 2005),
and choice satisfaction (Kim, Healey, Goldstein, Hasher, & Wipr-
zycka, 2008). In this article, we investigate whether age-related
changes in both cognitive capacities and affect impact decisions in
a sequential decision-making task.
Cognitive limitations may be one source of differences between
younger and older adults’ sequential decisions. Burns, Lee, and
Vickers (2006) showed that performance in the secretary problem
was correlated with fluid cognitive abilities. Aging has been con-
nected to decreases in such fluid abilities (Salthouse, 1996), and
limitations in fluid abilities may underlie age differences in pre-
decisional information search (Mata & Nunes, 2010). For example,
Mata et al. (2007) found that older adults searched about 15% less
information before making a decision compared to younger adults
and that individual differences in fluid abilities could account for
age-related differences in search in a multiattribute decision task.
Bettina von Helversen and Rui Mata, Department of Psychology, Uni-
versity of Basel and Max Planck Institute for Human Development, Berlin,
Germany.
This work was supported by a research grant by the German Research
Foundation to Bettina von Helversen (RI 1226/5). We thank Hung Quach and
Gregor Caregnato for assistance in preparing and conducting the experiment,
Ryan Murphy for help with the optimal solution for the sequential decision
making task, and Susanne Scheibe for helpful comments on this work.
Correspondence concerning this article should be addressed to Bettina
von Helversen, University of Basel, Department of Psychology, Mission-
sstraße 62a, CH-4055 Basel. Switzerland. E-mail: bettina.vonhelversen@
unibas.ch
Psychology and Aging © 2012 American Psychological Association
2012, Vol. ●●, No. , 000– 000 0882-7974/12/$12.00 DOI: 10.1037/a0027845
1
In sum, to the extent that sequential decisions tax individuals’
cognitive abilities, there is potential for age differences in these
decision tasks.
Affective experience is another likely source of differences
between younger and older adults’ sequential decisions. Aging is
associated with higher emotional competence (Blanchard-Fields,
2007; John & Gross, 2004; Phillips, Henry, Hosie, & Milne, 2008;
Scheibe & Blanchard-Fields, 2009) and improved affective expe-
rience (Carstensen, 2006; Carstensen et al., 2011; Charles, Reyn-
olds, & Gatz, 2001; Röcke, Li, & Smith, 2009). Affect may
influence peoples’ decision making by affecting people’s search
tendencies: Higher positive affect has been connected to less
information search in judgments (Fiedler, Renn, & Kareev, 2010),
consumer decisions (Beatty & Ferrell, 1998), multiattribute deci-
sion tasks (Isen & Means, 1983), and sequential decision tasks
(von Helversen et al., 2011). Positive affect may reduce search by
generally promoting superficial thinking (see Bless & Fiedler,
2006; Schwarz & Clore, 2007). In addition, people in a positive
mood tend to evaluate attractive objects even more positively (e.g.,
Adaval, 2003; Bower, 1991; Howard & Barry, 1994), which in a
sequential decision-making task could translate into terminating
search early by accepting an object that would otherwise be
rejected.
In sum, past work suggests that systematic age differences in
both affect and cognitive abilities may lead to older adults search-
ing less relative to younger adults prior to making a decision and
may, ultimately, impact decision quality in sequential decision
making. The present study aims to test these scenarios by asking
younger and older adults to make decisions in a sequential
decision-making task.
Study 1
We asked younger and older adults to perform a sequential
decision-making task in the form of a computerized shopping task.
Participants were asked to shop for 60 different consumer products
(e.g., LCD monitors, lawn mowers, refrigerators) with the goal of
finding the lowest priced offer. For each product (e.g., LCD
monitor), participants could see up to 40 offers varying only in
price. They had to decide when to stop search and accept the
current offer. Participants were also regularly asked to indicate
their performance goals as well as their satisfaction with their
performance. Thus, the shopping task uses an everyday scenario to
assess how younger and older adults differ in search behavior,
goals, and choice satisfaction, as well as whether these translate
into differential decision quality.
Our expectation was that older adults would search fewer op-
tions (Mata & Nunes, 2010) and be equally or potentially more
satisfied with their choices relative to younger adults (Kim et al.,
2008). We also asked participants to complete affect and cognitive
ability measures because we aimed to test the hypotheses that
positive affect and/or cognitive limitations are linked to search in
sequential choice. We thus hoped to understand the contribution of
affect and cognitive abilities to age differences in sequential de-
cision making.
Method
Participants. Sixty-four people, 32 younger adults (18
women, mean age !24.2, SD !2.7) and 32 older adults (17
women; mean age !69.0, SD !3.5) participated in the study.
Younger adults were students from one of the Berlin universities
(mean years of education !15.6, SD !4.6). Older adults were
community-dwelling adults (mean years of education !15.3,
SD !2.9), recruited from the participant database of the Max
Planck Institute for Human Development, Berlin, Germany. Par-
ticipation took between 1 and 2 hr and participants received on
average 18 for their participation. The study was approved by the
Ethics Committee of the Max Planck Institute for Human Devel-
opment.
Affect and cognitive measures. Mean values for affect and
cognitive measures are provided in Table 1. Affect was measured
with the German version (Krohne, Egloff, Kohlmann, & Tausch,
1996) of the Positive and Negative Affect Schedule (PANAS;
Watson, Clark, & Tellegen, 1988), consisting of 10 positive and 10
negative affective words, such as excited or distressed. Participants
rated how well each item described their current mood on a scale
from 1 (not at all)to7(very much). Positive and negative affect
scores were calculated by taking the mean rating of the positive
and negative items, respectively. Reliability was adequate for both
scales and both measurement occasions (all Cronbachs’ "#.88).
Participants also completed a number of cognitive tasks, namely, a
vocabulary test (Lehrl, 1999), a processing speed task (digit-
symbol substitution; Wechsler, 1981), and the cognitive reflection
test (Frederick, 2005). The cognitive reflection test is a three-item
measure (e.g., “a bat and a ball cost $1.10 in total; the bat costs
$1.00 more than the ball; how much does the ball cost?”) and is
thought to measure one’s ability to engage in effortful inference
processes and avoid judgment biases (Frederick, 2005; Oechssler,
Roider, & Schmitz, 2009).
Shopping task problems. The participants’ task was to pur-
chase consumer products (e.g., LCD monitors, lawn mowers,
refrigerators) for the lowest price possible. The prices for the
different consumer products shown to participants were realistic so
as to maximize the likelihood that older adults would remember
these (Castel, 2005). For this purpose, we searched for the lowest
and the highest price for each product on Internet websites and
then generated prices by drawing values from a normal distribution
with a mean equal to the average value of the highest and lowest
prices and a standard deviation set so that 98% of the prices would
fall between the highest and the lowest price.
Procedure. Participants first completed the affect measure.
Participants then read the instructions for the shopping task and
completed a practice trial. For the practice trial and each of the
following 60 consumer products (e.g., LCD monitor), participants
could search through 40 price offers that were presented sequen-
tially in a random order. At each step, an offer was presented, and
participants could choose to accept or reject the offer at their own
pace. Additionally, participants were informed about the number
of remaining offers for a specific product (see Figure 1 for a screen
shot). If the offer was rejected, it expired and participants were
presented with the next offer. An expired offer could not be chosen
at a later point of time. If participants had not accepted an offer for
a product (e.g., LCD monitor) before they had seen all 40 offers
they were forced to accept the last offer. If an offer was accepted,
the product (e.g., LCD monitor) was bought for the offered price,
and participants received explicit feedback about its rank and the
points earned. Then participants moved on to the next product
(e.g., lawn mower). Participants were paid according to the rank of
2VON HELVERSEN AND MATA
the selected offer. Rank refers to the relative price of the selected
offer compared to the 40 offers for that product, the cheapest
offer has a rank of 1, the second cheapest a rank of 2, and so on.
Participants received 40 points for the best offer (i.e., Rank 1),
39 for the second-best offer (Rank 2), and so on. At the end of
the experiment points were converted to at a rate of 100
points !.5 .
The 60 products were aggregated in 12 blocks consisting of five
products each. In each of the 12 blocks, participants were first
asked to indicate their performance goals by indicating the Desired
Rank of the offers that they aspired to reach. Specifically, they had
to indicate how high (i.e., from 1 to 40) the offers they selected
should be ranked to satisfy them with their performance. They then
completed the five games corresponding to five different consumer
products, and finally indicated their Satisfaction with their perfor-
mance on a 5-point scale (1 !not satisfied at all,5!very
satisfied). After completing the shopping task, participants again
completed the affect measure, the measures of cognitive ability,
and a number of questionnaires that are not the focus of this
article.
1
The data on measures of cognitive abilities for four
participants (two younger and two older) were lost.
Results
Performance and search. Performance and search are key
behavioral measures in the sequential choice task. Performance
was measured as the average rank of the options participants
selected across the 60 problems encountered. Search was measured
as the average number of options considered before making a
choice. Mean values for younger and older adults are presented in
Table 1. Descriptive analyses for each age group indicated two
outliers that performed three standard deviations below their peers
and that we excluded from further analyses:
2
One younger adult
selected options with an average rank of 11 compared to an
average rank of 5 for younger adults, and one older adult chose
options with an average rank of 22 compared to an average rank of
7 for older adults. These numbers also indicate that older adults
1
Additionally, participants filled out the Satisficing and Maximizing
questionnaire by Schwartz et al. (2002) and the Preference for Intuition and
Deliberation questionnaire by Betsch (2004).
2
The results do not substantially change when considering all partici-
pants.
Table 1
Participant Characteristics, Decision Task Variables, and Threshold Parameters by Age Group in Study 1
Measures
Younger adults
(n!31)
Older adults
(n!31) Statistical test
M SD M SD t (df) p d
Participant characteristics
Vocabulary 29.83 4.57 34.00 2.75 4.21 (56) .001 1.11
Processing speed 66.17 14.75 46.66 11.82 5.56 (56) .001 1.46
CRT 1.24 1.12 0.83 1.04 1.46 (56) .15 0.38
Positive affect 1 4.09 0.96 5.20 0.83 5.95 (60) .001 1.24
Negative affect 1 1.77 0.79 1.49 0.66 3.55 (60) .001 0.38
Positive affect 2 4.21 0.96 5.12 1.07 1.50 (60) .14 0.90
Negative affect 2 1.41 0.64 1.31 0.52 0.64 (60) .53 0.17
Decision task
Performance 5.18 0.90 6.61 1.87 3.83 (43.19) .001 0.97
Search 18.61 4.20 15.90 6.09 2.04 (53.29) .05 0.52
Goals 4.60 2.23 2.71 1.24 4.13 (46.95) .001 1.05
Satisfaction 3.31 0.57 3.16 0.58 1.07 (60) .29 0.26
Multiple threshold model
Threshold 1 9.00 3.54 6.61 5.52 2.02 (60) .05 0.52
Threshold 2 20.03 3.36 16.90 6.94 2.23 (43.34) .03 0.57
Threshold 3 25.42 4.58 21.55 6.66 2.66 (60) .01 0.68
Error 0.42 0.23 0.69 0.38 3.38 (60) .001 0.86
Note. Sample size for vocabulary, processing speed, and cognitive reflection was 29 for both younger adults and older adults because the data concerning
these measures for four participants was lost. For performance, search, goals, and threshold 2, we report corrected degrees of freedom (df) for the ttest,
because a Levene test indicated unequal variances. CRT !cognitive reflection test.
Figure 1. Screenshot of the shopping task.
3
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
performed substantially worse than younger adults: On average,
older adults selected offers about two ranks below those selected
by younger adults (see Table 1 for statistical tests). To better gauge
participants’ performance, we considered three reference points:
an optimal strategy, a multiple-threshold strategy, and a random
strategy. The optimal strategy possesses knowledge about the
quality distribution of the offers and uses a decreasing threshold
based on the expected value that is updated at each decision step.
The optimal strategy would allow selecting the second-best option
out of 40 on average. A decision maker without distribution
knowledge could rely on a multiple-threshold strategy that approx-
imates optimal performance in this task (Bearden et al., 2006; von
Helversen et al., 2011) and select the fourth-best option out of 40
on average. In turn, random choice would lead to selecting the
20th-best option out of 40 on average. When contrasting partici-
pants’ performance to these benchmarks, one can conclude that
although both age groups performed on average worse than an
optimal strategy, both younger and older adults did clearly better
than chance. In sum, the multiple-threshold strategy matches most
closely participants’ average performance. In the following sec-
tion, we provide evidence that such a strategy indeed captures
participants choices well.
Next, we compared the average numbers of objects searched by
younger and older adults. We found that older adults searched
overall fewer offers compared to younger adults (see Table 1 for
statistical tests). In the secretary problem one would expect that
search length has a quadratic relationship to performance—
searching too little as well as too much can lead to suboptimal
performance. As illustrated in Figure 2 (left panel), we found in an
analysis including both age groups that search had a quadratic
relation to performance, explaining 57% of the variance in perfor-
mance, F(2, 59) !38.32, p!.001. Separate analyses for older
and younger adults showed similar results, with search explaining
35% of the variance in performance in younger adults, F(2, 28) !
7.48, p$.01, and 55% of the variance in performance in older
adults, F(2, 28) !17.16, p$.01.
Modeling behavior in the secretary problem. To better
understand how younger and older participants solved the sequen-
tial choice task, we computationally modeled their choices with a
multiple-threshold strategy. The multiple-threshold model has
been successfully applied to sequential decision-making tasks with
rank-dependent payoffs (Bearden et al., 2006; von Helversen &
Johnson, 2008; von Helversen et al., 2011). The model has several
parameters that can be interpreted as capturing internal thresholds
for accepting an offer that is best, second best, and so on, in
comparison with previously seen offers. For example, if the first
threshold is 12, the participant would not accept any of the first 11
offers, but would accept, from offer 12 onward, the next offer that
was better than any of the previous offers. In the same vein, the
second threshold captures from when on in the search, a participant
would accept an offer that is only the second-best of the offers he
or she has seen thus far. We assumed that participants’ choices are
in line with a multiple-threshold strategy, but that the parameter
values of the thresholds can differ between participants. To find
the threshold values that best explained the participants’ choices,
we estimated the best fitting threshold parameter values to the data
of individual participants choosing the threshold values that max-
imized the number of choices predicted by the model. Additionally
we implemented an error parameter to capture choices that devi-
ated from model predictions because the participant rejected the
first option that was predicted by the model. This error parameter
may capture unintentional rejections, for instance, because a par-
ticipant clicked too fast through the options. Alternatively, the
error parameter can be interpreted as variance in choice, implying
that thresholds are not deterministic but probabilistic in nature.
A model relying on three threshold parameters captured partic-
ipants’ choices well, explaining 68% of all choices (for more
details on the multiple-threshold model and model fitting see
Appendix A). In comparison, a baseline model using a single
parameter that predicts that the same number of options is searched
in all trials could explain only 12% of participants’ choices. Older
and younger adults were equally well described, suggesting that
010203040
2
4
6
8
10
12
14
Number of Options Searched
Perf orma nce (Mea n Ra nk)
010203040
1
2
3
4
5
6
7
Number of Options Searched
Posi tive A ffect
Older Adults
Yo u n g e r A d u l t s
Figure 2. Search, performance, and positive affect. The left panel shows the curvilinear relation between search
and performance (lower values indicate higher ranks) for younger and older adults; the right panel shows the
linear relation between search and positive affect separately for younger and older adults. Markers denote age
group: younger adults, black; older adults, gray.
4VON HELVERSEN AND MATA
the model can capture the decisions of both age groups equiva-
lently. As can be seen in Table 1, in accordance with the behav-
ioral results on search, older adults had lower thresholds than
younger adults, suggesting that older adults searched less than
younger adults because they had lower thresholds for accepting an
option.
Affect, cognitive ability, and sequential decision making.
Can differences in affect or cognitive ability account for why older
adults searched fewer options and had less stringent thresholds
relative to younger adults? Affect was measured before and after
the task. Older adults reported higher positive but similar negative
affect relative to younger adults at both time points (see Röcke et
al., 2009, for a similar result). Concerning the cognitive measures,
older adults performed worse on the processing speed task, simi-
larly on the cognitive reflection task, and better in the vocabulary
task relative to younger adults (see Table 1 for statistical tests).
We relied on correlation analyses to investigate the link between
both affect and cognitive abilities and sequential decision making.
Regarding affect, we focused on the first measurement, as affect
after the task could be influenced by task performance. As can be
seen in Table 2, higher positive affect was related to worse
performance, less search, and lower threshold parameters. To
investigate how specific emotional states related to sequential
choice behavior, we additionally analyzed correlations between the
decision variables and individual affect items. Overall, search and
performance were related to the majority of affect items, but the
item enthusiastic correlated highest with both search, r(62) !
%.36, p$.01, and Threshold 1, r(62) !%.38, p$.01. Complete
reports of the item level correlations can be found in Appendix B
(Table B1). Negative affect was not related to performance or
search. Regarding the cognitive measures, processing speed was
negatively related to performance, but none of the cognitive mea-
sures was significantly related to search length or thresholds. The
results were supported by an additional regression analysis on
search, with positive and negative affect and cognitive abilities as
predictors. Only positive affect emerged as a significant predictor,
b!%.38, t(52) !2.56, p!.01; including age group in the
regression reduced the impact of positive affect somewhat, b!
%.32, t(51) !1.92, p!.06. The effect of age group was no longer
significant, b
age
!%.16, t(51) !%.78, p!.44. In sum, the results
seem to suggest that individual differences in positive affect but
not fluid cognitive abilities were related to search in the sequential
decision task. To find out if the relation would hold for both
younger and older adults, we also conducted similar analyses
within each age group (see Table 3): We found no relation between
cognitive or positive affect measures and sequential decision mak-
ing in the younger sample. In contrast, in older adults, positive
affect showed a strong relation to search (see Figure 2, right
panel), and similar albeit weaker correlations with threshold pa-
rameters. In sum, our results suggest that positive affect, but not
cognitive ability, is related to reduced search in older adults.
Affect, performance goals, and choice satisfaction. Older
adults performed worse and reported higher performance goals
relative to younger adults, yet they showed levels of satisfaction
with their choices similar to those of younger participants (see
Table 1). These findings raise the possibility that older adults’
choice satisfaction may not be simply a function of objective
decision performance but may depend on other factors such as
positive affect. Across all participants, performance was related to
choice satisfaction, in the direction that higher satisfaction was
related to better performance, r(62) !%.27, p!.03. Overall,
positive affect did not correlate with satisfaction, r(62) !.08, p!
.55. However, when we analyzed whether positive affect was
correlated with choice satisfaction separately for the two age
groups, we found a marginal correlation for older adults: r(31) !
.32, p!.08, albeit not for younger adults: r(31) !.05, p!.76.
This result was supported by an additional multilevel analysis
taking advantage of the repeated measures of satisfaction and
performance. The analysis showed that whereas satisfaction was
related to performance for older and younger adults, positive affect
was only related to satisfaction for older adults (for details on the
analysis see Appendix C). Overall, these findings suggest that
positive affect, at least in older adults, can influence subjective
aspects of decision making, such as choice satisfaction.
Discussion of Study 1
We investigated how younger and older adults solved a sequen-
tial decision-making task. Overall, older adults considered fewer
options and choose worse options than younger adults. For both
younger and older adults, search was closely related to perfor-
mance, suggesting that searching less may have contributed to age
differences in decision performance. We also modeled partici-
pants’ decision processes with a formal model that assumes that
decision makers use multiple thresholds to decide when to buy an
option. The model provided a good fit to younger and older adults’
Table 2
Correlations Between Sequential Decision-Making Variables, Cognitive Measures, and Affect
Speed
Cognitive
reflection Vocabulary Positive affect Negative affect
Performance !.35 %.25 .22 .37 %.06
Search .05 %.06 %.02 !.40 .13
Threshold 1 .08 .02 .02 !.31 .18
Threshold 2 .05 .04 .04 !.26 .07
Threshold 3 .06 %.06 %.01 !.25 .11
Error %.12 %.11 .08 .26 %.18
Note. Total sample size was 62, but sample size for the correlations involving cognitive ability measures
(speed, cognitive reflection, vocabulary) is 58 due to lost data. Coefficients significantly different from zero at
the .05 level are in boldface. Positive and negative affect refer to the measures taken before the sequential choice
task.
5
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
choices and estimates for the threshold parameters corroborated
the behavioral results on search: Older adults had significantly
lower values on threshold parameters relative to younger adults,
suggesting that older adults were willing to accept options earlier
in the search process. These findings parallel others suggesting that
older adults prefer having fewer options to choose from (Reed,
Mikels, & Simon, 2008), tend to make immediate decisions
(Meyer, Talbot, & Ranalli, 2007), and search less information
prior to making a decision in multiattribute decision tasks (Mata &
Nunes, 2010).
We also aimed to assess the contribution of affect and cognitive
abilities to age differences in search behavior in sequential deci-
sion making. Correlation analyses suggested that positive affect
but not cognitive abilities, such as speed of processing, were
related to search in the sequential decision task. The result supports
the notion that increased positive affect can influence peoples’
decision making by inducing people to search less either through
promoting superficial thinking (e.g., Bless & Fiedler, 2006;
Schwarz & Clore, 2007) or overly positive evaluations of options
(e.g., Adaval, 2003; Bower, 1991; Howard & Barry, 1994). A
closer look at the data revealed, however, that positive affect was
related to search in older but not younger adults.
There are two possible explanations for our finding of a link
between positive affect and decision making in older but not
younger adults. First, the relation between positive affect and
lower thresholds for accepting an option is specific to older adults
in line with claims that older adults rely more on affect when
making decisions than younger adults (e.g., Hanoch, Wood, &
Rice, 2007). Second, higher levels of positive affect decrease
thresholds for accepting an option in both younger and older
adults, but in our study younger adults’ naturally occurring indi-
vidual differences in affect were not sufficient to impact decision
behavior. The second explanation implies that younger adults
should behave more like older adults and thus tend to accept
options earlier whenever they experience higher levels of positive
affect. We conducted a second study in which we manipulated
positive affect in younger adults to test whether higher levels of
positive affect lead to accepting an option earlier in sequential
decisions.
Study 2
We manipulated mood in younger adults to investigate if ele-
vated positive affect would lead to choice behavior similar to that
of older adults in the sequential choice task. The study had two
conditions: a positive affect condition and a neutral affect condi-
tion.
Method
Participants. Eighty-one students from the University of
Basel participated in Study 2 (40 in the neutral and 41 in the
positive condition; M
age
!23.42, SD !6.11; 86% women).
Participants received course credit or a show-up fee and earned
between 3 and 7 Swiss Francs additionally depending on their
performance in the task.
Design and procedure. Participants completed the same
sequential choice task as described in Study 1. Affect was manip-
ulated prior to completing the sequential decision-making task.
The affect manipulation consisted of showing participants 15
pictures from the International Affective Picture System (IAPS;
Lang, Bradley, & Cuthbert, 2008) for 7 s each prior to the decision
task (for similar mood manipulations see Dreisbach & Goschke,
2004; Pin˜o´n & Ga¨rling, 2004). Additionally, we showed partici-
pants four pictures after every five trials of the decision task. The
pictures were selected based on the ratings of valence and arousal.
In the positive condition, participants saw pictures that had re-
ceived highly positive ratings. In the neutral condition, participants
saw pictures that had received average ratings. Arousal was kept
constant between the conditions. Because ratings of valence and
Table 3
Correlations Between Sequential Decision-Making Variables, Cognitive Measures, and Affect
Separately for Younger and Older Adults
Speed
Cognitive
reflection Vocabulary Positive affect Negative affect
Younger adults
Performance %.06 .02 %.04 .17 %.01
Search .07 %.12 %.001 %.05 .20
Threshold 1 .12 .04 .08 %.003 .37
Threshold 2 %.08 .01 .27 .15 .12
Threshold 3 %.21 %.26 .06 %.03 .28
Error .15 %.09 %.27 %.02 %.29
Older adults
Performance %.16 %.31 .04 .14 .05
Search %.23 %.09 .23 !.46 %.01
Threshold 1 %.17 %.05 .21 %.31 %.03
Threshold 2 %.17 %.02 .19 %.22 %.14
Threshold 3 %.08 %.03 .28 %.09 .01
Error .16 %.02 %.04 .05 %.06
Note. Sample size for the correlations involving cognitive ability measures (Speed, cognitive reflection,
vocabulary) is 29 for both younger and older adults due to lost data. Coefficients significantly different from zero
at the .05 level are in boldface. Positive and negative affect refer to the measures taken before the sequential
choice task.
6VON HELVERSEN AND MATA
arousal differ by gender (Fessler, Pillsworth, & Flamson, 2004;
Lang et al., 2008), we selected different sets of pictures for men
and women to equate the impact of the pictures. The average
ratings of valence and arousal for the selected pictures by condi-
tion and gender are reported in Table 4. Affect was measured with
the PANAS (Krohne et al., 1996; Watson et al., 1988) at three time
points: before the affect manipulation (Time Point 1), after the
affect manipulation and before the sequential choice task (Time
Point 2), and after the sequential choice task (Time Point 3).
Results
Manipulation of affect. Participants showed similar levels of
initial positive affect in both conditions, t(79) !0.49, p!.63, d!
0.10 (for means and standard deviations see Table 5). After the
affect manipulation, positive affect increased in the positive con-
dition, but was stable in the neutral condition, resulting in higher
positive affect in the positive than in the neutral condition, t(79) !
2.28, p!.02, d!0.50. During the task, positive affect decreased
in both conditions. At Time Point 3, participants in the positive
condition reported marginally larger levels of positive affect than
the neutral condition, t(79) !1.83, p!.07, d!0.41. Negative
affect differed between conditions, with participants in the positive
condition reporting lower levels of negative affect than partici-
pants in the neutral condition at Time Point 1, t(79) !2.71, p!
.01, d!0.60, and Time Point 2, t(79) !3.02, p!.003, d!0.66,
but not at Time Point 3, t(79) !1.47, p!.15, d!0.30. Overall,
these results suggest that the manipulation of affect was successful
in that it increased positive affect differences between the two
groups. In addition, the two groups also differed initially in neg-
ative affect by chance (participants were allocated randomly to the
two conditions), and this difference remained significant after the
positive mood induction. Consequently, the two groups differed in
both positive and negative affect, which made it possible for us to
investigate the role of both on sequential decision making.
Affect and sequential choices. To describe behavior in the
sequential choice task, we again measured performance as the
average rank of the selected options and search as the average
number of offers considered. As in Study 1 we also modeled
participants’ behavior with the multiple-threshold model. The
model described participants’ choice well, explaining 70% of their
choices, SD !9.71 (see Table 6 for means and standard devia-
tions).
To analyze whether the affect manipulation influenced behavior
in the sequential choice task, we compared participants’ behavior
in the neutral and the positive conditions. We did not find differ-
ences between the conditions for search or performance (for sta-
tistical tests, see Table 6). However, participants in the positive
condition had significantly lower values on the first threshold
parameter than participants in the neutral condition, t(79) !2.33,
p!.02.
We conducted correlation analyses to investigate the role of
positive and negative affect on threshold parameters estimated
from the computational model of sequential decision making.
Positive affect immediately after the mood manipulation was neg-
atively correlated with the first threshold parameter, r(81) !%.21,
p!.055. This correlation was of similar magnitude when con-
trolling for negative affect at the first two time points, partial
r(77) !%.21, p!.067, suggesting that the effect is independent
of negative affect. Analyses conducted on individual affect items
revealed that the first threshold was specifically correlated to items
measuring positive valence, such as excited,r(81) !%.26, p!
.02, and enthusiastic,r(81) !%.29, p$.01, but not to items
measuring attentiveness, such as alert,r(81) !%.06, p!.58, or
attentive,r(81) !.003, p!.98. For the complete item level
correlations, see Appendix B (Table B2).
Regarding negative affect, we found a negative correlation
between negative affect immediately after the mood manipulation
and the second threshold parameter, suggesting that higher nega-
tive affect led to lower thresholds for second-best offers, r(81) !
%.25, p!02. This correlation was of similar magnitude when
controlling for positive affect at the first two time points, partial
r(77) !%.25, p!.03, suggesting that the effect is independent of
positive affect. An analysis on the item level showed that this
correlation was driven by the items upset,r(81) !%.31, p$.01,
nervous:r(81) !%.23, p!.04, and jittery:r(81) !%.26,
p!.02.
Finally, participants in the positive and the neutral condition did
not differ in their performance goals or reported satisfaction (see
Table 6). Neither performance goals nor satisfaction correlated
with positive or negative affect (all ps#.27). Satisfaction was also
not correlated with search or performance in the sequential choice
task.
Discussion of Study 2
We conducted an affect manipulation that increased younger
adults’ positive affect in a positive relative to a neutral condition.
Table 4
Valence and Arousal Measures of the Selected Affective Stimuli
Measures
Affect condition
Neutral Positive
Men Women Men Women
Valence 5.16 (0.23) 5.09 (0.28) 7.27 (0.30) 7.95 (0.33)
Arousal 4.29 (0.48) 4.31 (0.47) 4.68 (0.50) 4.65 (0.50)
Note. Values are means with standard deviations in parentheses.
Table 5
Positive and Negative Affect in Study 2
Affect condition
Neutral (N!40) Positive (N!41)
PA
1
4.25 (0.98) 4.34 (0.83)
PA
2
4.17 (1.02) 4.67 (0.96)
PA
3
3.87 (1.17) 4.36 (1.23)
NA
1
1.77 (0.56) 1.44 (0.54)
NA
2
1.61 (0.61) 1.24 (0.50)
NA
3
1.43 (0.48) 1.27 (0.50)
Note. Values are means with standard deviations in parentheses. PA
1
/
NA
1
!positive/negative affect before the affect manipulation; PA
2
/NA
2
!
positive/negative affect after the affect manipulation and before the se-
quential choice task; PA
3
/NA
3
!positive/negative affect after the sequen-
tial choice task.
7
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
Although participants in the two conditions did not differ signifi-
cantly in overall search length or performance, the first threshold
for accepting an offer estimated from computational modeling
was, as expected, lower in the positive relative to the neutral
condition. The results suggest that high positive affect may in-
crease the likelihood of accepting an attractive offer and thus that
affect can impact younger adults’ sequential decision making.
Analyses of affect at the item level suggested that the link
between acceptance thresholds and positive affect was particularly
strong for items measuring positive valence, as opposed to atten-
tiveness. According to Watson and Clark (1994) positive affect
encompasses more specific emotional states such as joviality,
self-assurance, and attentiveness: Joviality consists of items fo-
cused on the valence of the affective state, whereas attentiveness
captures alertness or energy. These results suggest that an overall
positive affect score such as that used in the PANAS may be too
general to capture the effects of positive valence on sequential
decision making in younger adults.
Our results also indicate that negative affect was related to
thresholds for accepting options. Higher negative affect and spe-
cifically higher ratings of being upset, nervous, and jittery were
related to accepting second-best options early. The role of negative
affect on search and thresholds is not straightforward. Negative
affect is comprised of specific emotional states such as fear, anger,
or sadness (Watson & Clark, 1994), and these emotional states
have been shown to have opposing influences on decision behav-
ior. Whereas sadness and fear have been associated with elaborate
processing and increased search (Bless & Fiedler, 2006; French,
Hevey, Sutton, Kinmonth, & Marteau, 2006; Lerner, Gonzalez,
Small, & Fischhoff, 2003; von Helversen et al., 2011), anger is
usually associated with reduced information processing (Fessler et
al., 2004; Lerner et al., 2003; Lerner & Keltner, 2001). Thus,
although our findings suggest that negative affect may lower
thresholds in sequential decision making, these results should be
interpreted cautiously as we did not manipulate specific negative
emotions. Gender effects may further complicate the pattern of
effects regarding negative emotions: Men and women differ in
how emotions such as anger or disgust influence their decisions
(e.g., Fessler et al., 2004). We did not find evidence that men and
women were differentially influenced by the mood induction or
differed significantly in their search behavior in Study 2. However,
gender effects may have been masked by the skewed gender ratio
in our sample, and so it would be important to examine the role of
negative emotions in a more gender-balanced sample.
In sum, the second study showed that a manipulation leading to
increased positive affect can lead to lower thresholds to accept an
attractive option in a sample of younger adults. Consequently, the
combined results of Study 1 and 2 support the idea that affect can
have an impact on sequential search in both older and younger
adults.
General Discussion
We investigated how cognitive abilities and affect influence
older and younger adults’ sequential decision making. In Study 1,
we found that older adults performed worse than younger adults
possibly because of reduced search. In addition, our results suggest
that positive affect, but not fluid abilities, contributed to older
adults searching less than younger adults: Older adults reported
overall higher levels of positive affect, and older adults’ positive
affect was related to search length and lower acceptance thresh-
olds. Study 1 did not find a relation between younger adults’
positive affect and any of the decision variables, raising the pos-
sibility that the link between affect and decision making is unique
to the older group. Study 2 manipulated affect in a sample of
younger adults to test whether affect can affect sequential decision
processes in younger adults, and the results showed that higher
levels of positive affect lowered the first acceptance threshold.
Taken together, this suggests that affect plays an important role in
sequential choice regardless of age group but that naturally occur-
ring differences in affect between younger and older adults may
contribute to age differences in sequential decision making.
Affect and Search
We found that positive affect was related to acceptance thresh-
olds in older adults (Study 1) and younger adults (Study 2). One
explanation for this result is that positive affect leads to more
superficial processing (e.g., Bless & Fiedler, 2006; Schwarz &
Clore, 2007). Alternatively, high levels of positive affect may
Table 6
Decision Task Variables by Affect Condition in Study 2
Measures
Affect condition Statistical test
Neutral
a
Positive
a
t(79) pd
Decision task
Search 16.67 (4.91) 15.98 (3.76) 0.71 .48 0.15
Performance (rank) 5.62 (1.23) 5.61 (1.20) 0.03 .98 0.01
Performance goals 4.39 (2.34) 4.65 (2.36) 0.51 .61 0.11
Satisfaction 3.33 (0.58) 3.21 (0.67) 0.89 .38 0.19
Multiple threshold model
Threshold 1 7.58 (3.51) 6.05 (2.28) 2.33 .02 0.52
Threshold 2 16.05 (5.62) 16.93 (5.35) 0.72 .47 0.16
Threshold 3 21.73 (4.80) 23.49 (5.36) 1.56 .12 0.35
Error 0.55 (0.21) 0.60 (0.25) 0.95 .34 0.22
Note. N
neutral
!40; N
positive
!41.
a
Values are means with standard deviations in parentheses.
8VON HELVERSEN AND MATA
increase the perceived attractiveness of options, thus increasing the
likelihood that one is selected early on (e.g., Adaval, 2003; Bower,
1991; Howard & Barry, 1994).
Our results are suggestive of links between affect and search in
sequential decision tasks but raise at least two outstanding issues
regarding this link. First, our results suggest that more theorizing
must be done regarding the role of different facets of positive
emotion in younger and older adults. For example, our results
suggest that there may be effects of valence but not attentiveness
on search in younger adults (Study 2), but we observed a link
between all positive affect items and search in older adults (Study
1). Second, the results from the two studies are at odds regarding
the role of negative affect: Although we found no effects of
negative affect in Study 1, there were effects of negative affect on
the second threshold parameter in Study 2. Because we relied on
natural variation in affect in Study 1 and did not manipulate
specific positive or negative emotions in Study 2, we cannot draw
strong inferences from our results. Future work that considers
more specific affect manipulations could prove useful in disentan-
gling the role of different affect dimensions and specific emotions
on search in sequential choice.
Affect and Satisfaction
Older and younger participants differed markedly in perfor-
mance, yet we did not find differences between older and younger
adults’ choice satisfaction. This result is particularly striking given
that older adults had more ambitious performance goals than their
younger counterparts. The effects may be partly explained by the
high levels of positive affect reported by older adults, which were
related to high levels of satisfaction and possibly protected older
adults from experiencing dissatisfaction after not reaching their
performance goals. Kim et al. (2008) found that older adults
reported improved choice satisfaction relative to younger adults if
given the opportunity to justify their choices. One potentially
interesting line of research would be to assess whether justification
processes underlie the resilience of older adults’ choice satisfac-
tion in the face of unmet goals and negative performance feedback.
Tolerable choice satisfaction in the face of poor performance
can be problematic if it prevents older adults from improving their
performance even when provided with negative performance feed-
back, as was the case in our task. Future work should aim to test
manipulations that can improve older adults’ decision in this
context, for example, by providing relative feedback regarding the
savings of other, perhaps younger participants.
Limitations and Future Directions
Decline in cognitive abilities has been suggested as the principal
factor underlying age-related differences in performance in
decision-making tasks, impairing the ability to seek and evaluate
information necessary for making a decision (Finucane, Mertz,
Slovic, & Scholze Schmidt, 2005; Henninger, Madden, & Huettel,
2010; Mata et al., 2007; Mata et al., 2010; Sharit, Hernandez,
Czaja, & Pirolli, 2008). Thus one would expect performance in
sequential decision making to be related to cognitive abilities
(Burns et al., 2006). In contrast, we found only a small relation
between processing speed and performance and no relation be-
tween cognitive abilities and overall search or acceptance thresh-
olds. One possible reason for this null finding is that search in our
sequential task was not sufficiently taxing on participants’ cogni-
tive abilities. Alternatively, our measures may have not been
sensitive enough to capture age differences in the relevant cogni-
tive abilities. For example, we did not measure working memory
capacity, which has been identified as a key factor contributing to
age differences in decision tasks (e.g., Mata et al., 2007) and
correlates substantially with fluid abilities that have been found to
predict performance in the secretary problem (Burns et al., 2006;
Kane & Engle, 2002). Future research should expand the measures
of cognitive abilities to allow quantifying their contribution to age
differences in performance in sequential choice.
We compared older and younger adults’ behavior in a laboratory
task that was modeled on Internet shopping where options are
frequently evaluated sequentially. Older adults are less familiar
with using Internet sites and thus lack of familiarity with such a
setting could have contributed to age differences in performance
(Sharit et al., 2008; Sharit, Hernandez, Nair, Kuhn, & Czaja,
2011). In addition, although we selected a broad range of products
and matched prices to actual offers on the Internet, older and
younger adults could have differed in their knowledge of product
prices, which in turn may have influenced their willingness to
accept an offer. Future work should control for experience with
sequential choice tasks and knowledge of price distributions to
assess the role of past experience to age differences in sequential
decision making.
Our results may have implications for real world decision mak-
ing. We found that older adults performed substantially worse than
younger adults in a laboratory decision-making task, and to the
extent that our task mimics natural settings, one could expect older
adults to show decreased decision quality in real-world choices
involving sequential evaluation of options. Nevertheless, there are
reasons to believe that reduced search may not always lead to poor
decision outcomes. For example, Mata and Nunes (2010) found
that using less information has a negligible effect on decision
quality in consumer decisions in which options were presented
simultaneously. Likewise, some conditions may foster competent
decisions in the face of limited search in sequential decisions; for
example, limited search may have negligible effects when options
differ little in quality and/or are presented in decreasing order of
quality. Consequently, a description of the real-world environ-
ments faced by younger and older adults is crucial to understand to
what extent age differences in search lead to poorer decision
outcomes.
Conclusion
Positive affect is generally considered a good thing. Increased
positive affect has been linked to health and longevity (Pressman
& Cohen, 2005), success (Lyubomirsky, King, & Diener, 2005),
and enhanced creativity and problem solving (Estrada, Isen, &
Young, 1997; Isen & Labroo, 2003). However, positive affect has
a darker side when it leads to superficial or stereotypical thinking
(Bless & Fiedler, 2006; Schwarz & Clore, 2007). Our results
suggest that high levels of positive affect as reported by elderly
persons can lead to insufficient search in sequential decision
making. Our results are thus compatible with the view that positive
affect may have costs for older adults’ decision making.
9
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
References
Adaval, R. (2003). How good gets better and bad gets worse: Understand-
ing the impact of affect on evaluations of known brands. Journal of
Consumer Research, 30, 352–367. doi:10.1086/378614
Bearden, J. N., Rapoport, A., & Murphy, R. O. (2006). Experimental
studies of sequential selection and assignment with relative ranks. Jour-
nal of Behavioral Decision Making, 19, 229 –250. doi:10.1002/bdm.521
Beatty, S., & Ferrell, M. E. (1998). Impulse buying: Modeling its precur-
sors. Journal of Retailing, 74, 169 –167. doi:10.1016/S0022-4359(98)
90009-4
Betsch, C. (2004). Pra¨ ferenz fu¨r Intuition und Deliberation. Inventar zur
Erfassung von affekt- und kognitionsbasiertem Entscheiden. [Preference
for Intuition and Deliberation (PID): An Inventory for Assessing Affect-
and Cognition-Based Decision-Making]. Zeitschrift fu¨ r Differentielle
und Diagnostische Psychologie, 25, 179 –197. doi:10.1024/0170-1789
.25.4.179
Blanchard-Fields, F. (2007). Everyday problem solving and emotion: An
adult developmental perspective. Current Directions in Psychological
Science, 16, 26 –31. doi:10.1111/j.1467-8721.2007.00469.x
Bless, H., & Fiedler, K. (2006). Mood and the regulation of information
processing and behaviour. In J. P. Forgas (Ed.), Affect in social thinking
and behaviour (pp. 65– 84). New York, NY: Psychology Press.
Bower, G. (1991). Mood congruity of social judgments. In J. P. Forgas
(Ed.), Emotion and social judgments. International series in experimen-
tal social psychology (pp. 31–53). Elmsford, NY: Pergamon Press.
Bruine de Bruin, W., Parker, A. M., & Fischhoff, B. (2010). Explaining
adult age differences in decision-making competence. Journal of Behav-
ioral Decision Making. Advance online publication. doi:10.1002/
bdm.712
Burns, N. R., Lee, M. D., & Vickers, D. (2006). Are individual differences
in performance on perceptual and cognitive optimization problems de-
termined by general intelligence? Journal of Problem Solving, 1, 5–19.
Retrieved from http://docs.lib.purdue.edu/jps/vol1/iss1/3
Carstensen, L. L. (2006). The influence of a sense of time on human devel-
opment.*** Science, 312, 1913–1915. doi:10.1126/science.1127488
Carstensen, L. L., Turan, B., Scheibe, S., Ram, N., Ersner-Hershfield, H.,
Samanez-Larkin, G. R.,...Nestleroade, J. R. (2011). Emotional expe-
rience improves with age: Evidence based on over 10 years of experi-
ence sampling. Psychology and Aging, 26, 21–33. doi:10.1037/
a0021285
Castel, A. D. (2005). Memory for grocery prices in younger and older
adults: The role of schematic support. Psychology and Aging, 20, 718 –
721. doi:10.1037/0882-7974.20.4.718
Charles, S. T., Reynolds, C. A, & Gatz, M. (2001). Age-related differences
and change in positive and negative affect over 23 years. Journal of
Personality and Social Psychology, 80, 136 –151. doi:10.1037/0022-
3514.80.1.136
Dreisbach, G., & Goschke, T. (2004). How positive affect modulates
cognitive control: Reduced perseveration at the cost of increased dis-
tractibility. Journal of Experimental Psychology Learning, Memory, and
Cognition, 30, 343–353. doi:10.1037/0278-7393.30.2.343
Estrada, C., Isen, A., & Young, M. (1997). Positive affect facilitates
integration of information and decreases anchoring in reasoning among
physicians. Organizational Behavior and Human Decision Processes,
72, 117–135. doi:10.1006/obhd.1997.2734
Ferguson, T. S. (1989). Who solved the secretary problem? Statistical
Science, 4, 289 –296. doi:10.1214/ss/1177012493
Fessler, D. M. T., Pillsworth, E. G., & Flamson, T. J. (2004). Angry men
and disgusted women: An evolutionary approach to the influence of
emotions on risk taking. Organizational Behavior and Human Decision
Processes, 95, 107–123. doi:10.1016/j.obhdp.2004.06.006
Fiedler, K., Renn, S.-Y., & Kareev, Y. (2010). Mood and judgments based
on sequential sampling. Journal of Behavioral Decision Making, 23,
483– 495. doi:10.1002/bdm.669
Finucane, M. L., & Gullion, C. M. (2010). Developing a tool for measuring
the decision-making competence of older adults. Psychology and Aging,
25, 271–288. doi:10.1037/a0019106
Finucane, M. L., Mertz, C. K., Slovic, P., & Schmidt, E. S. (2005). Task
complexity and older adults’ decision-making competence. Psychology
and Aging, 20, 71– 84. doi:10.1037/0882-7974.20.1.71
Frederick, S. (2005). Cognitive reflection and decision making. Journal of
Economic Perspectives, 19, 25– 42. Retrieved from http://www.jstor
.org/stable/4134953
French, D. P., Hevey, D., Sutton, S., Kinmonth, A. L., & Marteau, T. M.
(2006). Personal and social comparison information about health risk:
Reaction to information and information search. Journal of Health
Psychology, 11, 497–510. doi:10.1177/1359105306063324
Gilbert, J. P., & Mosteller, F. (1966). Recognizing the maximum of a
sequence. Journal of the American Statistical Association, 61, 35–73.
doi:10.2307/2283044
Hanoch, Y., Wood, S., & Rice, T. (2007). Bounded rationality, emotions
and older adult decision making: Not so fast and yet so frugal. Human
Development, 50, 333–358. doi:10.1159/000109835
Henninger, D. E., Madden, D. J., & Huettel, S. A. (2010). Processing speed
and memory mediate age-related differences in decision making. Psy-
chology and Aging, 25, 262–270. doi:10.1037/a0019096
Howard, D. J., & Barry, T. E. (1994). The role of thematic congruence
between a mood-inducing event and an advertised product in determin-
ing the effects of mood on brand attitudes. Journal of Consumer Psy-
chology, 3, 1–27. doi:10.1016/S1057-7408(08)80026-5
Isen, A., & Labroo, A. (2003). Some ways in which positive affect
facilitates decision making and judgment. In S. L. Schneider & J.
Shanteau (Eds.), Emerging perspectives on judgment and decision re-
search (pp. 365–393). Cambridge, UK: Cambridge University Press.
Isen, A., & Means, B. (1983). The influence of positive affect on decision-
making strategy. Social Cognition, 2, 18 –31. doi:10.1521/soco.1983
.2.1.18
John, O. P., & Gross, J. J. (2004). Healthy and unhealthy emotion regu-
lation: Personality processes, individual differences, and life span de-
velopment. Journal of Personality, 72, 1301–1334. doi:10.1111/j.1467-
6494.2004.00298.x
Kane, M. J., & Engle, R. W. (2002). The role of prefrontal cortex in
working-memory capacity, executive attention, and general fluid intel-
ligence: An individual-differences perspective. Psychonomic Bulletin &
Review, 9, 637– 671. doi:10.3758/BF03196323
Kim, S., Healey, M., Goldstein, D., Hasher, L., & Wiprzycka, U. J. (2008).
Age differences in choice satisfaction: A positivity effect in decision
making. Psychology and Aging, 23, 33–38. doi:10.1037/0882-7974
.23.1.33
Krohne, H., Egloff, B., Kohlmann, C., & Tausch, A. (1996). PANAS
Positive and Negative Affect Schedule—Deutsche Fassung. Diagnos-
tica, 42, 139 –156.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008). International
affective picture system (IAPS): Affective ratings of pictures and instruc-
tion manual (Technical Report A-8). Gainesville, FL: University of
Florida.
Lehrl, S. (1999). Mehrfachwahl-Wortschatz-Intelligenztest: Manual mit
Block MWT-B. Balingen, Germany: Spitta.
Lerner, J. S., Gonzalez, R. M., Small, D. A., & Fischhoff, B. (2003).
Effects of fear and anger on perceived risks of terrorism: A national field
experiment. Psychological Science, 14, 144 –150. doi:10.1111/1467-
9280.01433
Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of
Personality and Social Psychology, 81, 146 –159. doi:10.1037/0022-
3514.81.1.146
Lewandowsky, S., & Farell, S. (2010). Computational modeling in cogni-
tion: Principles and practice. London, UK: Sage Publications.
Löckenhoff, C. E., & Carstensen, L. L. (2007). Aging, emotion, and
10 VON HELVERSEN AND MATA
health-related decision strategies: Motivational manipulations can re-
duce age differences. Psychology and Aging, 22, 134 –146. doi:10.1037/
0882-7974.22.1.134
Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent
positive affect: Does happiness lead to success? Psychological Bulletin,
131, 803– 855. doi:10.1037/0033-2909.131.6.803
Mata, R., & Nunes, L. (2010). When less is enough: Cognitive aging,
information search, and decision quality in consumer choice. Psychology
and Aging, 25, 289 –298. doi:10.1037/a0017927
Mata, R., Schooler, L., & Rieskamp, J. (2007). The aging decision maker:
Cognitive aging and the adaptive selection of decision strategies. Psy-
chology and Aging, 22, 796 – 810. doi:10.1037/0882-7974.22.4.796
Mata, R., von Helversen, B., Karlsson, L., & Cu¨ pper, L. (2011). Adult age
differences in categorization and multiple-cue judgment. Developmental
Psychology. Online advance publication. doi:10.1037/a0026084
Mata, R., von Helversen, B., & Rieskamp, J. (2010). Learning to choose:
Cognitive aging and strategy selection learning in decision making.
Psychology and Aging, 25, 299 –309. doi:10.1037/a0018923
Mather, M., Knight, M., & McCaffrey, M. (2005). The allure of the
alignable: Younger and older adults’ false memories of choice features.
Journal of Experimental Psychology: General, 134, 38 –51. doi:10.1037/
0096-3445.134.1.38
Meyer, B. J. F., Talbot, A. P., & Ranalli, C. (2007). Why older adults make
more immediate treatment decisions about cancer than younger adults.
Psychology and Aging, 22, 505–524. doi:10.1037/0882-7974.22.3.505
Oechssler, J., Roider, A., & Schmitz, P. W. (2009). Cognitive abilities and
behavioral biases. Journal of Economic Behavior & Organization, 72,
147–152. doi:10.1016/j.jebo.2009.04.018
Phillips, L. H., Henry, J. D., Hosie, J. A., & Milne, A. B. (2008). Effective
regulation of the experience and expression of negative affect in old age.
Journals of Gerontology: Series B: Psychological Sciences, 63B, P138 –
P145. doi:10.1093/geronb/63.3.P138
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., & R. Core Team. (2009).
nlme: Linear and nonlinear mixed effects models. R package version
3.1–92. Retrieved from http://CRAN.R-project.org/package!nlme
Pin˜o´n, A., & Ga¨ rling, T. (2004). Effects of mood on adoption of loss frame
in risky choice. Göteborg Psychological Reports, 34, 1–11.
Pressman, S. D., & Cohen, S. (2005). Does positive affect influence health?
Psychological Bulletin, 131, 925–971. doi:10.1037/0033-2909.131
.6.925
R Development Core Team. (2009). R: A language and environment for
statistical computing. Vienna, Austria: R Foundation for Statistical
Computing. Retrieved from http://www.R-project.org
Reed, A. E., Mikels, J. A., & Simon, K. I. (2008). Older adults prefer less
choice than younger adults. Psychology and Aging, 23, 671– 675. doi:
10.1037/a0012772
Röcke, C., Li, S. C., & Smith, J. (2009). Intraindividual variability in
positive and negative affect over 45 days: Do older adults fluctuate less
than young adults? Psychology and Aging, 24, 863– 878. doi:10.1037/
a0016276
Salthouse, T. A. (1996). The processing-speed theory of adult age differ-
ences in cognition. Psychological Review, 103, 403– 428. doi:10.1037/
0033-295X.103.3.403
Scheibe, S., & Blanchard-Fields, F. (2009). Effects of emotion regulation
on cognitive performance: What is costly for young adults is not so
costly for older adults. Psychology and Aging, 24, 217–223. doi:
10.1037/a0013807
Schwartz, B., Ward, A., Monterosso, J., Lyubomirsky, S., White, K., &
Lehman, D. R. (2002). Maximizing versus satisficing: Happiness is a
matter of choice. Journal of Personality and Social Psychology, 83,
1178 –1197. doi:10.1037/0022-3514.83.5.1178
Schwarz, N., & Clore, G. L. (2007). Feelings and phenomenal experiences.
In E. T. Higgins & A. Kruglanski (Eds.), Social psychology: Handbook
of basic principles (2nd ed., pp. 385– 407). New York, NY: Guilford
Press.
Seale, D. A., & Rapoport, A. (1997). Sequential decision making with
relative ranks: An experimental investigation of the secretary problem.
Organizational Behavior and Human Decision Processes, 69, 221–236.
doi:10.1006/obhd.1997.2683
Seale, D. A., & Rapoport, A. (2000). Optimal stopping behavior with
relative ranks: The secretary problem with unknown population size.
Journal of Behavioral Decision Making, 13, 391– 411. doi:10.1002/
1099-0771(200010/12)13:4$391::AID-BDM359#3.0.CO;2-I
Sharit, J., Hernandez, M. A., Czaja, S. J., & Pirolli, P. (2008). Investigating
the roles of knowledge and cognitive abilities in older adult information
seeking on the web. ACM Transactions on Computer-Human Interac-
tion, 15, 1–25. doi:10.1145/1352782.1352785
Sharit, J., Hernandez, M. A., Nair, S. N., Kuhn, T., & Czaja, S. J. (2011).
Health problem solving by older persons using a complex government
web site: Analysis and implications for web design. ACM Transactions
on Accessible Computing, 3, 1–35. doi:10.1145/1952383.1952386
von Helversen, B., & Johnson, T. P. (2008). Der Einfluss von “Satisficing”
und “Maximizing” auf das Entscheidungsverhalten. In W. Sarges & D.
Scheffer (Eds.), Innovative Ansa¨ tze fu¨r die Eignungsdiagnostik. Reihe:
Psychologie fu¨ r das Personalmanagement (pp. 265–273). Göttingen,
Germany: Hogrefe.
von Helversen, B., Wilke, A., Johnson, T., Schmid, G., & Klapp, B. (2011).
Performance benefits of depression: Sequential decision making in a
healthy sample and a clinically depressed sample. Journal of Abnormal
Psychology, 120, 962–968. doi:10.1037/a0023238
Watson, D., & Clark, L. A. (1994). The PANAS-X Manual for the Positive
and Negative Affect Schedule—Expanded Form. Ames, IA: The Univer-
sity of Iowa.
Watson, D., Clark, L., & Tellegen, A. (1988). Development and validation
of brief measures of positive and negative affect: The PANAS scales.
Journal of Personality and Social Psychology, 54, 1063–1070. doi:
10.1037/0022-3514.54.6.1063
Wechsler, D. (1981). Wechsler adult intelligence scale: Revised manual
(WAIS-R). New York, NY: Psychological Corporation.
Yeo, A. J., & Yeo, G. F. (1994). Selecting satisfactory secretaries. Aus-
tralian Journal of Statistics, 36, 185–198. doi:10.1111/j.1467-842X
.1994.tb00861.x
Zwick, R., Rapoport, A., Lo, A. K. C., & Muthukrishnan, A. V. (2003).
Consumer sequential search: Not enough or too much? Marketing Sci-
ence, 22, 503–519. doi:10.1287/mksc.22.4.503.24909
(Appendices follow)
11
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
Appendix A
Description of the Multiple-Threshold Strategy and Computational Modeling
To better understand how participants solved the sequential
choice task, we constructed a computational model to study par-
ticipants’ choices. In the original secretary problem, in which
payoffs are only received when the best option is found, Seale and
Rapoport (1997, 2000) showed that a single-threshold model pro-
vides the best description of participant behavior. In secretary
games with rank-dependent payoffs as in the task we use here, an
extension of the single-threshold strategy that uses multiple thresh-
olds is necessary to describe behavior (Bearden et al., 2006; von
Helversen, et al., 2011; Yeo & Yeo, 1994).
The multiple-threshold strategy assumes that participants set
thresholds that determine when they will accept an option with a
given relative rank. Relative rank refers to the rank of an option
compared with the options seen thus far. Accordingly if an option
is better than all options seen thus far, it has a relative rank of 1;
if it is better than all but one of the offers seen thus far, it has a
relative rank of 2 an so on. These thresholds exist for each relative
rank, meaning that participants set a threshold for each relative
rank that will halt search once an option that is, respectively, best,
second-best, third-best, and nth-best of the options seen so far is
encountered. For instance, if the threshold for an option with a
relative rank of 1 is five, from the sixth offer on the strategy will
accept any offer that is better than the first five offers. Because the
threshold for a relative rank of 2 is higher than the threshold for a
relative rank of 1 and the threshold for a relative rank of 3 is higher
than the threshold for a relative rank of 2 and so on, this strategy
will accept worse options as fewer options are left to choose from.
Figure A1 illustrates the thresholds of an optimal multiple-
threshold strategy that maximizes the average payoff in a task with
40 options and rank-dependent payoff. A person relying on the
optimal multiple-threshold strategy would not chose any option
before seeing option number 12, but from option number 12 on,
the participant would accept any option with a relative rank of 1
(that is an option that is better than all options seen thus far). From
option number 20 on, the person would accept any option with a
relative rank of 1 or 2; from option 26 on, any option with a
relative rank of 3 or lower would be chosen, and so on. The
optimal multiple-threshold strategy results in an average perfor-
mance of 37 points in our task, which corresponds to selecting the
fourth best option out of 40.
The multiple-threshold strategy has been found to describe
human behavior well, although participants generally have lower
thresholds than predicted by the optimal model (e.g., Bearden et
al., 2006; von Helversen & Johnson, 2008). We assumed that
participants’ choices are in line with a multiple-threshold strategy,
but that the parameter values of the thresholds can differ between
participants. To find the threshold values that best explained the
participants’ choices, we estimated the best-fitting parameter val-
ues to the data of individual participants. We only estimated the
best-fitting values for thresholds for relative ranks of 3 or lower.
We did not try to estimate further thresholds because in our task
the majority of the choices, 88%, fell on an option with a relative
rank of 3 or lower and only 12% on options with a relative rank of
4 or higher. Thus, there were not enough data to obtain stable
estimates for parameter values for thresholds for a relative rank of
4 or higher. Please note that the model with three thresholds
encompasses simpler versions of the model with only one or two
thresholds.
We found the best-fitting threshold values for each participant
by implementing a grid search in Matlab. More specifically, we
calculated for each participant the number of choices that was
predicted by any possible combination of threshold values, with
each threshold taking a value between 0 and 40 (see Lewandowsky
& Farell, 2010, for an overview on model fitting). We then chose
the threshold values that maximized the number of choices pre-
dicted by the model. Additionally, we implemented an error pa-
rameter to capture choices that deviated from model predictions
because the participant rejected the first option that was predicted
by the model, but then chose the next option the model predicted.
This error parameter may capture unintentional rejections for in-
stance because a participant clicked too fast through the options.
Alternatively, the error parameter can be interpreted as variance in
choice, implying that thresholds are not deterministic but proba-
bilistic in nature.
(Appendices continue)
010 20 30 40
0
5
10
15
20
25
30
35
40
Number of options seen
Maximal relative rank of option
Figure A1. The maximal relative rank an option may have to be accepted
according to the optimal multiple-threshold strategy at each point of the
game.
12 VON HELVERSEN AND MATA
We also considered two baseline models: a random choice
model and a simple search model, which assumes that participants
always search through the same amount of options. A model
assuming random choice predicts on average 4% of participants’
choices. The simple search model was able to predict 12% of
participants’ choices. In comparison, the multiple-threshold strat-
egy with the estimated threshold values predicted participants’
responses very well, explaining 68% (SD !11.06) of participants’
choices and 78% (SD !12.11) of choices of options with a
relative rank of 3 or lower, suggesting it is a good model to
describe participants’ behavior. In addition, older and younger
adults, where equally well fit by the model, M
young
!77.4, SD !
12.51, M
old
!78.39, SD !11.88, t(60) !.32, p!.75. Average
parameter values can be found in Table 1.
Appendix B
Correlations Between Behavioral Measures in the Sequential Choice Task and Items
Measuring Affect
Table B1
Correlations Between Affect Items at Time 1 and Sequential Choice Measures in Study 1
Search Rank Threshold 1 Threshold 2 Threshold 3
Positive affect
PA scale !.40
!!
.37
!!
!.31
!
!.26
!
!.25
!
Interested !.30
!
.37
!!
%.14 %.18 %.24
Excited !.26
!
.19 !.32
!
%.24 %.19
Strong %.09 .03 %.05 %.09 %.11
Enthusiastic !.35
!!
.25
!
!.38
!!
%.23 %.14
Proud %.16 .20 %.23 !.27
!
%.11
Alert %.24 .14 %.06 %.01 %.23
Inspired !.30
!
.36
!!
%.21 %.17 %.07
Determined !.31
!
.27
!
%.16 %.14 %.13
Attentive !.29
!
.28
!
%.19 %.16 !.26
!
Active !.30
!
.36
!!
%.22 %.19 %.25
Negative affect
NA scale .13 %.06 .18 .07 .11
Distressed .16 %.16 .22 .19 .27
!
Upset .04 %.05 .09 .03 .16
Guilty .11 %.06 .06 %.02 %.03
Scared %.05 %.02 %.05 %.01 %.12
Hostile .16 %.19 .11 .07 .08
Irritable .26
!
%.21 .27
!
.13 .13
Ashamed .09 %.01 .14 .07 .13
Nervous .00 .18 .07 %.09 .01
Jittery .12 %.02 .17 .06 .09
Afraid .08 %.08 .17 .07 .04
Note. N !62; PA !positive affect; NA !negative affect. Coefficients significantly different from zero at the
.05 level are in boldface.
!
p$.05.
!!
p$.01.
(Appendices continue)
13
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
Table B2
Correlations Between Affect Items at Time 2 and Sequential Choice Measures in Study 2
Search Rank Threshold 1 Threshold 2 Threshold 3
Positive affect
PA scale %.12 .09 %.21 .03 .16
Interested %.05 %.03 %.12 .04 .09
Excited %.13 .07 !.26
!
.02 .06
Strong %.03 .13 %.17 .07 .13
Enthusiastic %.11 .06 !.29
!!
.12 .10
Proud %.12 .15 !.31
!!
%.04 .04
Alert %.06 .02 %.06 .03 .21
Inspired %.17 .07 %.13 %.09 .08
Determined %.08 .08 %.14 .01 .17
Attentive .00 .04 .00 .06 .22
!
Active %.05 .05 .00 .04 .10
Negative affect
NA scale %.12 .10 .08 !.25
!
%.18
Distressed %.06 .07 .08 %.18 %.11
Upset %.17 .31
!!
.00 !.31
!!
%.18
Guilty .03 %.04 .10 %.12 %.04
Scared %.14 .05 %.03 %.04 %.19
Hostile %.09 .15 .02 %.11 %.21
Irritable .01 .07 .13 %.13 %.10
Ashamed %.03 .08 .05 %.12 %.14
Nervous %.21 .08 %.01 !.23
!
%.08
Jittery %.08 .00 .10 !.26
!
%.11
Afraid %.09 %.01 .05 %.12 %.19
Note. N !81. PA !positive affect; NA !negative affect. Coefficients significantly different from zero at the
.05 level are in boldface.
!
p$.05.
!!
p$.01.
Appendix C
Multilevel Modeling of Choice Satisfaction
To make use of the repeated measurements of satisfaction and
performance from the sequential decision task, we used multilevel
modeling. We conducted an initial screening of the amount of
within-person variability in the satisfaction data using the intra-
class correlation obtained from an unconditional means model in
which the residual variance was significant. The analysis indicated
that about one third of the total variance in satisfaction was located
within persons (intraclass correlation !.31) and that group-mean
reliability was good (.85). We also conducted analyses separately
for the two age groups. The intraclass correlation was .27 and .36,
and group-mean reliability was .81 and .86, for younger and older
adults, respectively. We then used the following model to assess
the link between choice satisfaction, performance, and positive
affect (at Time Point 1):
Level 1: Satisfactionit !&0it "&1it(Performance) "rit
Level 2: &0i!'00 "'01(Positive Affect) "'02(Age Group) "u0i
&1i!'10
(Appendices continue)
14 VON HELVERSEN AND MATA
In Level 1, choice satisfaction of participant ion block tis a
function of the intercept (&
0it
), performance obtained in that block
(&
1it
; rank 1 to 40), and the residual (r
it
). In the Level 2 equations,
'
00
represents the mean satisfaction for younger adults (when
age !0), '
01
represents the effect of positive affect on mean
choice satisfaction, '
02
is the difference in average satisfaction
between younger and older adults (when age !1), '
10
captures the
effect of performance on satisfaction, and u
0i
and u
1i
are residuals.
We also tested additional models that considered whether the
effect of affect varied as a function of age group or whether
positive affect moderated the link between performance and sat-
isfaction, but these did not provide significantly better fits than the
simpler model described above. We grand-mean centered positive
affect and performance variables and estimated parameters for the
model using R (version 2.11.1; R Development Core Team, 2009),
and the nlme package (Pinheiro, Bates, DebRoy, Sarkar, & R Core
Team, 2009). As can be seen in Table C1, the parameter estimates
suggest that (a) performance was significantly related to satisfac-
tion, (b) younger and older adults did not differ significantly in
choice satisfaction, and (c) positive affect was correlated with
choice satisfaction. However, we also fitted similar models sepa-
rately for the two age groups (i.e., excluding the effect of age
group). As can be seen in Table C1, we found an effect of
performance for both age groups, but only older adults showed a
relation between positive affect and choice satisfaction, suggesting
that positive affect had a significant impact on satisfaction for
older but not younger adults.
Received July 28, 2011
Revision received February 17, 2012
Accepted February 21, 2012 "
Table C1
Results From the Multilevel Regression Models
Effects Coefficient SE df T p
All participants
Intercept ('
00
) 3.34 0.11 681 30.6 $.001
Performance ('
10
)%0.40 0.03 681 13.74 $.001
Positive affect ('
01
) 0.18 0.09 59 2.13 .04
Age group ('
02
)%0.20 0.17 59 1.18 .25
Younger adults
Intercept ('
00
) 3.26 0.13 340 25.46 $.001
Performance ('
10
)%0.47 0.05 340 10.06 $.001
Positive affect ('
01
) 0.08 0.14 29 0.58 .57
Older adults
Intercept ('
00
) 3.09 0.11 340 27.08 $.001
Performance ('
10
)%0.34 0.04 340 9.60 $.001
Positive affect ('
01
) 0.24 0.11 29 2.27 .03
15
AGE, AFFECT, AND SEQUENTIAL DECISION MAKING
... Some studies have investigated tasks closer to real sequential choice problems in which option values are drawn independently from a known distribution and the decision maker observes this exact values as each option is considered. (Guan et al., 2018;Guan et al., 2015;Kogut, 1990;Lee, 2006;von Helversen et al., 2012). In this version, the optimal solution is to choose according to a thresholds which is based on the probability of winning on the later positions (Gilbert et al., 1966, Section 3). ...
... If an option is better/worse than the threshold, it is accepted/rejected. However, there are differences in whether individuals have a single fixed threshold or the threshold decreases as the sequence progresses von Helversen et al., 2012). ...
... Based on the expected reward, a threshold T i is calculated for each day i (for a detailed description of the calculation of optimal thresholds, I refer to the Supporting Material of Chapter 2: A linear threshold model for optimal stopping behavior, Text A). Consequently, a price x i is accepted von Helversen et al., 2012) and suggest that people must rely on a different strategy when solving such tasks. I propose the linear threshold model (LTM, Baumann et al., 2020) which postulates that human's decision thresholds are constrained by a linear relation to each other. ...
... Past research has identified several factors that direct choices; short and long term gain (Marschner et al., 2005;Quartz, 2009), motivation for change (Renner et al., 2009), several cognitive abilities (e.g., processing speed; Missier et al., 2012), present goals (O'Donnell et al., 2017), potential risks, and affect (Lerner & Keltner, 2000;von Helversen & Mata, 2012) are some but not all of the factors that have been proposed in the past literature. Recently, mentally traveling to the personal past and future has begun to attract the attention of researchers for their role in decision-making processes (Bulley et al., 2016;De Luca et al., 2018;Suddendorf & Busby, 2005). ...
Article
Mentally traveling to one's personal past and future connects the current self with self at different points in time. When making decisions in the present, individuals benefit from their past and potential future decisions with the help of mental time travel. This review documents the theoretical and empirical studies concerning mental time travel to past and future decisions and their influence on current decision-making processes. Particularly, certain characteristics of past and future decisions directly or indirectly influence present life decisions. Synthesizing the existing literature, we developed a theoretical model suggests that temporally close decisions tend to include different type of details and diverge from distant decisions in terms of personal meaningfulness. Further, past and future decisions that are elaborated and meaningful, have a great impact on present decisions. Future empirical work testing this model has the potential to generate findings that will inform intervention strategies towards improved decision-making.
... The positivity effect has also been shown to further influence older adults in decision contexts, in that older relative to younger adults found positive healthrelated pamphlets to be more informative than negative pamphlets, and they also better remembered the positive versus negative messages (Shamaskin et al., 2010). Other work suggests that incidental positive affect may be harmful to the decision making of older versus younger adults, leading to more nonoptimal choices (Mikels et al., 2013) and reduced information search (von Helversen & Mata, 2012). Thus, this is yet another area where the emotional aspects of aging at work need to be accommodated to maintain and enhance both decision quality and motivation. ...
... Some studies have investigated tasks closer to real sequential choice problems by presenting the actual value of the option to the decision makers (6)(7)(8)(9)(10). In this version, the optimal is based on calculating the probability of winning on the later positions. ...
Article
In many real-life decisions, options are distributed in space and time, making it necessary to search sequentially through them, often without a chance to return to a rejected option. The optimal strategy in these tasks is to choose the first option that is above a threshold that depends on the current position in the sequence. The implicit decision-making strategies by humans vary but largely diverge from this optimal strategy. The reasons for this divergence remain unknown. We present a model of human stopping decisions in sequential decision-making tasks based on a linear threshold heuristic. The first two studies demonstrate that the linear threshold model accounts better for sequential decision making than existing models. Moreover, we show that the model accurately predicts participants’ search behavior in different environments. In the third study, we confirm that the model generalizes to a real-world problem, thus providing an important step toward understanding human sequential decision making.
Chapter
This chapter discusses age differences in complex decision making and judgment, particularly the role of motivational factors and individual differences. Literature on the influence of age-related changes in cognition and motivation on search and performance in complex decision making is reviewed. The role of financial incentives, need for cognition, and need for cognitive closure is discussed, including the age-related influence of motivational factors on the performance of sequential decision-making tasks. Additionally, the role of feedback as a factor producing superior performance of older adults in a decision-making task is introduced. Moreover, novel research findings regarding connections between intellectual helplessness and information and communication technologies in older adults are presented. Lastly, individual differences in numeracy and intellectual helplessness in mathematics as predictors of age-related differences in performance of multiattribute tasks are described.
Thesis
Développé en 2015, l’Emotion Imbued Choice (EIC ; Lerner et al., 2015) explique la manière dont les émotions influencent la prise de décision. Selon ce modèle, les émotions incidentes (non reliées à la décision) associées à un haut degré de certitude déclencheraient un traitement plutôt heuristique de l’information tandis que les émotions associées à un haut degré d’incertitude déclencheraient un traitement plutôt délibératif (Tiedens & Linton, 2001), expliquant des différences de performances dans les tâches de prise de décision.L’influence des émotions incidentes a été étudiée dans la prise de décision unique et dans la prise séquentielle de décisions (i.e., séries de décisions). Si l’EIC, fournit un cadre explicatif clair concernant la prise de décision unique, ses prédictions sont moins explicites quant à la prise séquentielle de décisions. Cette dernière présente la particularité que chaque décision est suivie d’un feedback (gain ou perte) ayant une valeur émotionnelle (positive vs. négative). Cette source d’influence émotionnelle intégrée (reliée à la décision) peut modifier les décisions suivantes. Des auteurs ont observé que le déclenchement d’un traitement de l’information plutôt heuristique conduit les participants à prendre des décisions plus avantageuses que le déclenchement d’un traitement de l’information plutôt délibératif (Bagneux et al. 2013 ; voir aussi Bollon & Bagneux, 2013). Toutefois, l’implication du traitement de la valeur émotionnelle des feedbacks dans l’obtention de ce pattern de résultats n’a, à ce jour, pas reçu de confirmation directe.Notre objectif était d’identifier la manière dont les émotions incidentes, en fonction de leur degré de certitude, modulent la prise séquentielle de décisions. Au travers d’une série de sept expériences, nous avons montré comment ces émotions incidentes interagissent avec l’influence émotionnelle intégrée des feedbacks. Ainsi, seul le déclenchement d’un traitement plutôt heuristique conduit les participants à moduler leurs décisions en fonction du type (positifs vs. négatifs) de feedbacks reçus. Ces résultats concernent les tâches de prise séquentielle de décisions ambigües et risquées. La modulation du pattern de prise de décision a été mise en évidence, de manière classique, lors de l’induction d’émotions incidentes négatives et, de manière plus originale, lors de l’induction d’émotions incidentes positives. En second lieu, l’analyse des résultats et de la littérature a mis en évidence la fiabilité médiocre de la mesure du degré de certitude. Une série de six autres études, s’inscrivant dans la méthodologie générale de construction d’échelles, a été réalisée afin d’en recherche les origines. Nous avons montré que cette faible fiabilité était la double conséquence d’une définition incomplète et d’une opérationnalisation imparfaite de la dimension de certitude. À un niveau théorique, nous proposons que l’ensemble de nos résultats supportent la nécessité d’étendre le modèle EIC à la prise séquentielle de décisions. Sur le plan méthodologique, nos résultats justifient l’intérêt de développer un nouvel outil de mesure du degré de certitude.
Chapter
Full-text available
In this chapter, we provide an overview of how a strategy perspective fruitfully contributes to our understanding of aging effects on cognitive functioning and brain activations. We review previous research showing that people use a wide variety of strategies to accomplish cognitive tasks and how strategy use evolves during aging. Although strategic variations are modulated by individual differences and experimental conditions, older adults have been found to use fewer strategies, to use the more demanding strategies less often, to select the most appropriate strategy on each problem less often, and to be less efficient when executing a given strategy than young adults. Adopting a strategy approach enables better characterization of age-related changes observed in brain activations during task completion and contributes to specify the mechanistic and functional significance of age-related changes in neural recruitments. Finally, we review recent evidence suggesting that cognitive control processes underlie age-related changes in strategy use.
Article
Cambridge Core - Cognition - The Cambridge Handbook of Cognitive Aging - edited by Ayanna K. Thomas
Article
Now days, understanding how people make purchasing decisions in digital era are of growing importance for researchers and marketers. Customer's reviews and ratings are important for better understanding of customers and improving services. Whereas, Consumer's reviews and ratings are available online for wide range of products and services. This Study investigated how average consumer ratings, product attributes, and consumer reviews influenced on the online purchasing decisions of buyers. In line with previous research, study found that buyers use all types of information on online purchasing; they clearly preferred products with better attributes and with higher average consumer ratings. It has been found that the buyer preferences can be changed for the higher-rated products, if the product is overridden by a rich negative or positive review.
Article
A German adaptation of the Positive and Negative Affect Schedule (PANAS) is presented, the factorial structure as well as the internal consistency are analyzed and external associations are described. A principal component analysis of state affect (N=349; instruction: "How do you feel at the moment") and trait affect(N=480; instruction: .,... In general") resulted in a distinct two-factor solution labeled "Positive Affectivity" (PA) and "Negative Affectivity" (NA). Applying four additional time-oriented instructions ("How did you feel today - during the past few days - weeks - the past yea"), an internal validation study indicated that as time intervals increase, the influence of the state affect decreases, whereas that of the trait affect increases. Trait PA and NA furthermore yielded differential associations with global personality traits and variables within the areas of anxiety, reporting of symptoms, and emotions, as well as stress management.
Article
Despite cognitive declines that occur with aging, older adults solve emotionally salient and interpersonal problems in more effective ways than young adults do. I review evidence suggesting that older adults (a) tailor their strategies to the contextual features of the problem and (b) effectively use a combination of instrumental and emotion-regulation strategies. I identify factors of problem-solving contexts that affect what types of problem-solving strategies will be effective. Finally, I discuss how this identification of factors affects what we know about developmental differences in everyday problem-solving competence.
Book
An accessible introduction to the principles of computational and mathematical modeling in psychology and cognitive science This practical and readable work provides students and researchers, who are new to cognitive modeling, with the background and core knowledge they need to interpret published reports, and develop and apply models of their own. The book is structured to help readers understand the logic of individual component techniques and their relationships to each other.
Article
We examine multi-period observation and selection problems with an unknown number of applicants in which applicants are interviewed one at a time on each period, recall of applicants that were interviewed and rejected is not possible, the decision on each period to reject or accept an applicant is based on relative ranks, and the objective is to maximize the probability of accepting the top-ranked applicant. We propose and then assess the efficiency of three descriptive models by simulation, and then test them competitively in a computer-controlled experiment. A cutoff decision model, in which the first r−1 applicants are rejected and then the first applicant who is ranked higher than all previously observed applicants is accepted, outperforms the two other two models. Compared with the optimal policy, subjects stop the search too early. Their behavior is accounted for by a cutoff model that postulates an endogenous cost of search. Copyright © 2000 John Wiley & Sons, Ltd.