Multitasking: productivity effects and gender differences

Thomas Buser Noemi Peter ∗

February 22, 2011

Abstract

We examine how multitasking affects performance and check whether women are indeed better
at multitasking. Subjects in our experiment perform two different tasks according to three treat-
ments: one where they perform the tasks sequentially, one where they are forced to multitask,
and one where they can freely organize their work. Subjects who are forced to multitask perform
significantly worse than those forced to work sequentially. Surprisingly, subjects who can freely
organize their own schedule also perform significantly worse. Finally, our results do not support
the stereotype that women are better at multitasking. Women suffer as much as men when forced
to multitask and are actually less inclined to multitask when being free to choose.

∗ Both authors are affiliated with the University of Amsterdam. We are grateful to Robert Dur, Hessel Oosterbeek,
Arthur Schram, Joep Sonnemans, Roel van Veldhuizen and seminar participants in Amsterdam for their comments and
suggestions. We gratefully acknowledge financial support from the University of Amsterdam through the Speerpunt Be-
havioural Economics and thank CREED (Center for Research in Experimental Economics and Political Decision-Making)
for letting us use their lab. Contact: [email protected], [email protected]. Website (Thomas Buser): buser.economists.nl.

1
1 Introduction
There are two common stereotypes about multitasking: juggling tasks is bad for productivity and
women are better at it. Surprisingly, both these questions remain scientifically underexplored. This
paper fills this gap through an experimental design which allows us to answer the following research
questions. First, how does multitasking affect productivity? Second, do people choose the optimal
amount of multitasking? Third, are there indeed gender differences in multitasking ability? And
fourth, are there gender differences in the propensity to multitask?1
The first pair of questions is motivated by a practical concern: how to schedule complex tasks
optimally? Is sequential execution advisable, or is it more productive to alternate? Is it optimal to let
workers choose their own schedule or should companies impose one? Although these questions are
important for both workers and companies, economists have traditionally ignored them. The classic
paper on multiple tasks (Holmstrom & Milgrom, 1991) is concerned not with scheduling but focuses
on asset ownership and incentive contracts between principals and agents.
Fortunately, the importance of work schedules recently gained some attention. Coviello et al.
(2010) show that judges who work on many court cases in parallel for presumably exogenous reasons
take more time than judges who work sequentially to complete similar portfolios of cases. Their
identification strategy uses random variation in the difficulty and urgency of cases (as judged by
other judges) and a procedural rule that the first hearing should be held no later than 60 days from
filing. They state that “individual productivity cannot be explained only in terms of effort, ability and
experience. Individual work scheduling (how much juggling is done) is a crucial input that cannot be
omitted from the production function of individual workers” (p. 2).
We share the view that work schedules are an important factor of productivity, but instead of
scheduling multiple projects, our research focuses on scheduling multiple tasks. Given that the latter
have a much shorter time horizon, we are able to use a lab experiment to examine our research
questions. This allows us to randomly allocate subjects to work schedules and thus to circumvent
the endogeneity issues that arise when using non-experimental data. Our subjects have to perform
two separate tasks according to one of three different treatments: one where they perform the tasks
sequentially, one where they are forced to alternate between the two tasks, and one where they can
freely organize their work. The amount of time spent on each task is identical in each treatment and
performance differences between treatments therefore measure the productivity effect of the different
schedules.
There is a related literature on ‘task-switching’ in psychology (see Monsell, 2003 for a review).
In these experiments, a series of stimuli is presented to participants who have to perform a short task
on each stimulus. For example, pairs of numbers are shown and subjects have to either add them up
or to multiply them (see Rubinstein et al., 2001). From time to time, the required operation changes.
1 For
the scope of the entire paper, by multitasking we mean switching back and forth between two cognitive tasks.
The concepts of multitasking and task-switching are discussed in more detail in Section 2.

2
It is commonly found that there are ‘switching costs’ associated with changing tasks, i.e. the response
to the stimuli is slower after a task-switch. This literature can, however, not answer our research
questions adequately. The tasks used are too simple to expect any advantages from multitasking
and - related to this problem - subjects are not allowed to choose their schedule freely.2 Also, these
experiments are not usually incentivized. In contrast, we use two complex tasks of longer duration
(a Sudoku and a Word Search puzzle). Therefore subjects can expect an advantage from alternating:
they can switch when they get stuck and later look at the same problem with a ‘fresh eye’. Indeed,
our subjects do switch when they are allowed to, enabling us to measure the effect of a self-chosen
work schedule as well.
Finally, none of the psychological experiments are designed to examine gender differences. More-
over, their samples are generally too small to do so and often characterised by strong gender imbal-
ances. Our comparatively large and balanced sample, on the other hand, allows us to test both whether
there are gender differences in multitasking ability and in the propensity to multitask. This pair of re-
search questions is motivated by the gap between popular views and scientific evidence: multi-million
selling books advertise that women are better at multitasking as a scientifically established fact3 , while
in reality this gender difference has not so far been shown by any peer-reviewed experimental paper.4
These views are so popular that some biological anthropologists attempt to theoretically explain the
empirically-not-yet-established phenomena. Fisher (1999), for example, claims that the prehistoric
division of work “built” different aptitudes into the male and female brain through natural selec-
tion. Different skills are required for hunting, performed by males, than for gathering, performed
by women. As a consequence, argues Fisher, women think “contextually”, as they synthesize many
factors into a “web of factors”, while men think linearly, focusing on a single task until it is done. Our
design enables us to test whether men indeed follow a more sequential schedule than women when
they are allowed to choose freely and whether they perform worse than women when they are forced
to multitask.
The paper proceeds as follows: Section 2 clarifies how do we define multitasking, the key concept
of the paper. Section 3 explains the details of the experimental design and describes the data. The
2 This is true even for papers that claim to examine voluntary task-switching. For example, Arrington & Logan (2004)
write that “We examined voluntary task switching by having subjects choose which task to perform on a series of bivalent
stimuli. Subjects performed parity or magnitude judgments on single digits. Instructions were to perform the two tasks
equally often and in a random order.” (p. 610, italics added)
3 See for example Pease & Pease (2001) and it’s adaptation, Why Men Can Only Do One Thing at a Time and Women

Never Stop Talking (Pease & Pease, 2003).

4 We searched extensively for peer-reviewed publications about gender differences in multitasking ability, but the

closest we could find is Criss (2006) and Havel (2004), which are manuscripts that are made available online at the
website of the National Undergraduate Research Clearinghouse. Both examined subjects who had to perform some
specified tasks while tallying keywords from a song/story. None of them found gender differences in productivity when
multitasking, but Criss (2006) found that women were better at accuracy. Nonetheless, we do not know whether the
findings can be attributed to multitasking as none of them had a control group. The media regularly mentions research
which supposedly shows that women are better at multitasking but to the best of our knowledge, none of this has been
published in peer-reviewed journals.

3
results are presented in Section 4 while their detailed discussion and the conclusions are presented in
Section 5.

2 Definitions
Multitasking is often thought of as the performance of multiple tasks at one time, but this definition
is at odds with the findings of many psychologists and neuroscientists. Pashler (1994) reviews the
related literature and concludes that our ability to simultaneously carry out even simple cognitive
operations is very limited. Using brain scanners, Dux et al. (2006) localize a neural network which
acts as a central bottleneck of information processing by precluding the selection of response to two
different tasks at the same time. Furthermore, Dux et al. (2009) show that while training can increase
the speed of information processing in this brain region, it remains true that the tasks are not processed
simultaneously but in rapid succession. Simultaneity is an illusion, which occurs if the tasks are so
simple that the alternations are very quick.5
Our definition of multitasking is consistent with the above findings: subjects working on two
cognitive tasks switch back and forth between them. Our design does not aim for resembling simul-
taneity but makes the alternation between tasks explicit. Typical situations of this kind are when an
employee’s work at hand is interrupted with another, perhaps more urgent task. Another example is
when people multitask on a computer, switching back and forth between windows or tabs.6
Note that our definition is similar to what psychologists call task-switching, but there is an im-
portant difference between the two: contingency. In our experiment, subjects continue working on
the same problem after they return from their work on the other task, so they can can benefit from
a ‘fresh eye’. In contrast, in the task-switching experiments subjects get a new stimulus to work on
each time (e.g. they get a new pair of numbers to add up, so only the operation remains the same, not
the problem they are working on).

3 Experimental design and data

3.1 Treatments and groups

Three treatments were applied during the experiment: Treatment Single, Treatment Multi, and Treat-
ment Choice (subjects were randomly allocated to treatments within each session and they did not
know these labels of course). In Treatment Single, subjects had to work on two tasks consecutively,
5 Simultaneity might be real if at least one of the tasks does not require information processing, but it seems hard to
find such a task.
6 The similarity with the latter is amplified by the fact that the experiment was conducted in a computer lab using a

web browser interface.

4
for 12 minutes each. In Treatment Multi, subjects were forced to switch between the two tasks approx-
imately every four minutes7 , resulting in the same total time constraint per task as before. Subjects
did not know how many switches would occur and the time intervals between switches varied, making
anticipation unlikely. In Treatment Choice, subjects could alternate between the two tasks by pressing
a ‘Switch’ button, subject to the same time constraint per task as before (12 minutes each). A timer
informed subjects about the remaining time for each task. When the 12 minutes for one task expired,
the screen changed automatically to the other task and the Switch button could not be used anymore.
It is important to see that this design ensures that the same amount of time is spent on each task
in all three treatments. If we tried to resemble simultaneity, for example by splitting the screen, we
could not determine how much time subjects spend on each task, and therefore we would not know
whether performance between treatments differs due to differences in the amount of time allocated to
the two tasks or due to differences in the schedules.
As shown in Table 1, subjects were assigned to three groups. Every subject played two rounds,
the first of which was Treatment Single. In the second round, subjects in Group 1 played Treatment
Single again, subjects in Group 2 played Treatment Multi, and subjects in Group 3 played Treatment
Choice. The subjects knew from the start that there would be two rounds and that they would work on
one Sudoku puzzle and one Word Search puzzle in each. The puzzles given in Round 2 were different
from the puzzles in Round 1 (but they were the same for all subjects within rounds).

Table 1: Treatments of each group

Group 1 Group 2 Group 3
Round 1 Single Single Single
Round 2 Single Multi Choice

This design allows us to answer all four research questions and the fact that Group 1 plays Single
twice allows for a difference-in-differences approach. This enables us to correct for learning effects
and performance drops due to exhaustion or boredom. To examine the effect of forced multitasking
on productivity, we can compare the performance difference between Round 1 and Round 2 of Group
2 to the performance difference of Group 1. To examine the effect of a self-chosen work schedule, we
can compare the performance-difference of Group 3 to the performance-differences of the other two
groups. If subjects choose the optimal work schedule, we should see that the performance-difference
of Group 3 is at least as high as the performance-difference of the other two groups.8 Note that
subjects already experienced an example of each task in Round 1, so we can assume that subjects in
Treatment Choice switch between tasks to maximize their payoff and not due to curiosity.
7 Gonzalez & Mark (2004) found that information workers spend on average 3 minutes on a task without interruption;
this average might be somewhat higher in a less fast-paced environment.
8 Since subjects in Group 3 can choose whether or not to alternate, finding that they performed worse than the other

groups would disprove that they chose optimally.

5
 Figure 1: Sudoku

To examine gender differences in multitasking ability, we follow a difference-in-difference-in-

differences approach. Note that any gender difference in performance can only come from differences
in the ability to multitask: since we compare performance in Round 2 to a subject’s own performance
in Round 1, performance differences cannot be led by differences in task proficiency. Besides, Group
1 captures any gender differences in learning or exhaustion.
Finally, to examine whether there is any gender difference in the propensity to multitask, we check
whether there is a gender difference in the number of switches in Treatment Choice. The propensity
to multitask might vary with proficiency: subjects who perform well might find switching easier or
more beneficial. Alternatively, subjects who get stuck more often may want to switch more often.
To avoid attributing such effects to gender differences in multitasking, we control for performance in
Round 1.

3.2 Tasks
Our design requires tasks that are not gender-specific and for which multitasking is natural and pos-
sibly beneficial. For these reasons, we have chosen Sudoku and Word Search as tasks. Sudoku is
played over a 9x9 grid, divided into 3x3 sub-grids called “regions”. The left panel of Figure 1 il-
lustrates that a Sudoku puzzle begins with some of the grid cells already filled with numbers. The
objective of Sudoku is to fill the other empty cells with integers from 1 to 9, such that each number
appears exactly once in each row, exactly once in each column, and exactly once in each region. The
numbers given at the beginning ensure that the Sudoku puzzle has a unique solution. For example,
the unique solution to the Sudoku in Figure 1 is illustrated the right panel. We measure performance
in the Sudoku task by the number of correctly filled cells.

6
 Figure 2: Word Search

When solving a Sudoku puzzle, solutions often come in waves. Multitasking can be appealing
when one is stuck: one can work on the other task and hope to see the problem from a different angle
when switching back.
The other task was to find as many words as possible in a Word Search puzzle. An example of
a Word Search puzzle is presented in the left panel of Figure 2, and its solution is presented in the
right panel. Participants had to look for the English names of European and American countries in
a 17x17 letter grid. Words could be in all directions, including diagonal and backwards. Subjects’
performance is measured by the number of correct words found.9
As in the case of Sudoku, it is reasonable to expect subjects to switch when unable to find new
words for a while. The situation is similar to polishing a paper, when reading the same lines over and
over becomes counterproductive after a while – one changes to another task simply because a ‘fresh
eye’ is needed to recognize meaning behind the letters.

3.3 Procedures, payments, timeline

One pilot and ten regular sessions were run in the computer lab of CREED (Center for Research in Ex-
perimental Economics and Political Decision-Making) at the University of Amsterdam. Participants
were university students from various fields of study. The application procedure ensured that the two
genders were represented approximately equally in every session, but left subjects unaware that the
experiment examines gender-related issues. The experiment was conducted in English, therefore both
9 Subjects did not know in advance how many words were hidden in the puzzle, but they knew that they would be
notified once all words were found.

7
international and Dutch students could participate. All instructions and tasks were computerized,10
and subjects were not allowed to use any paper or take notes during the experiment.
The experiment started with an introduction that explained the rules of the two tasks and gave the
participants opportunity to practice. Subjects learned that there would be two rounds and that they
would have to play a Sudoku and a Word Search in both rounds. In each round, subjects earned 6
points for each correctly filled Sudoku cell and lost 6 points for each cell filled with a wrong number
to avoid random guessing. Subjects were not penalized for cells filled with multiple numbers.11 They
received 9 points for each word found in Word Search. In Word Search, only entire words could be
marked and there was therefore no need to penalize random clicking. Subjects’ total points for each
round were determined as the sum of their points in Sudoku and their points in Word Search. Negative
total points were rounded up to 0. One of the two rounds was randomly selected for payment at the
end and the conversion rate was 1 euro per 11 points. In addition to this, there was a fixed show-up
fee of 7 euros. The performance payments and the conversion rate were chosen based on the results
of a pilot, such that subjects could earn approximately equal amounts on the two tasks and that the
average payment was around 23 euros. The sessions lasted for approximately 1 hour and 45 minutes.
The order of the tasks within each round was randomized, and the assignment of subjects to the
three treatments in round 2 was random as well, so that each group consisted of approximately one
third of the subjects in every session. The rules of the treatments were explained immediately before
the start of the treatment. Subjects were not aware of the fact that not everyone was playing the same
treatment as they did.
After both rounds were over, but before being informed about their payment, we elicited some
background information such as gender, age, field of study, and nationality from the subjects through
a questionnaire. Those who participated in Treatment Choice were also asked their reasons for (not)
switching.

3.4 Data
Our sample consists of 218 subjects from the ten regular sessions.12 They are 22 years old on av-
erage and the majority of them is Dutch (73 percent). Approximately half of the sample consists of
economics students (53 percent). The sample contains 11 censored observations from subjects who
solved the entire Sudoku puzzle in the second round but not in the first.13 As Section 3.1 explained,
10 The program was written in PHP (an HTML-embedded scripting language) and was displayed using the web browser
Mozilla Firefox.
11 Subjects could enter multiple numbers in one cell to denote uncertainty.
12 We only use the data from the regular sessions because some parameters were changed after the pilot.
13 In addition, 17 subjects solved the entire puzzle in the first round and 11 of them also in the second round. These 11

subjects are excluded since we do not know how their performance changed from the first to the second round. We also
dropped the six subjects who solved the puzzle only in the first round. Otherwise we would encounter a sample selection
problem: among the best performers of Round 1, we would only drop those who fall back enough in Round 2 to not solve
the entire puzzle. Inclusion in the sample is thus conditional on not having solved the entire Sudoku in Round 1. Recall

8
 Table 2: Number of observations per cell
Men Women Sample
Group 1 30 40 70
Group 2 39 31 70
Group 3 43 35 78
Total 112 106 218

Table 3: Average total points per cell

Group 1 (n=70) Group 2 (n=70) Group 3 (n=78)
Round 1 Round 2 Diff. Round 1 Round 2 Diff. Round 1 Round 2 Diff.
Men 184 188 4 186 172 -14 195 174 -22
Women 185 192 7 198 177 -21 205 198 -7
Both 185 190 6 191 174 -17 200 185 -15
Note: all numbers are rounded to the nearest integer.

subjects were randomly assigned to three groups. Table 2 shows the number of observations per group
and gender.14 As we can see, there are between 30 and 43 subjects per cell.

4 Results

4.1 Multitasking and performance

Performance is measured as the sum of Sudoku plus-points and Word Search points.15 Table 3 shows
means per group and gender (for both rounds), and performance differences between rounds. Note
that the difference-in-differences(-in-differences) strategy takes care of any performance-differences
between cells in Round 1.
Comparing the results of Group 1 and Group 2 to each other shows that the productivity-effects
of multitasking are significantly negative: the difference-in-differences is -23 points (ranksum test;
p-value=0.02). Subjects who could pick their own schedule (Group 3) perform only slightly better
than those forced to multitask and score 21 points less than Group 1 (ranksum test; p-value=0.06).
The difference-in-differences in performance between men and women in Group 2 suggests that
men handle forced multitasking relatively better than women, but the difference is not significant
(ranksum test; p-value=0.84). The results of Group 3, on the other hand, suggest that women are
better at organizing their own schedule, but this difference is not significant either (ranksum test; p-
value=0.45). There are no gender differences in learning either: the performance improvement for
that every subject receives treatment Single in Round 1; therefore inclusion is independent of treatment. The proficiency
of the dropped subjects is remarkably different from the rest of the sample: they solved all 51 Sudoku-cells in the first
round, while the average was 20 cells (with a standard deviation of 11), and the second best performance was 42 cells.
14 The distribution of the dropped subjects is as follows: 5 from Group 1, 8 from Group 2 and 4 from Group 3.
15 Sudoku minus-points were only used to discourage random guessing, not to measure performance.

9
Group 1 subjects is the same for both genders (ranksum test; p-value=0.87). In sum, simple non-
parametric rank-sum tests do not detect any significant gender differences.
Using regression techniques, we can check whether the results hold if we take censoring and
the (non-significant) gender differences in learning into account. Table 4 shows the results of fixed
effects and first-difference censored regressions which take full advantage of the panel structure of
our data.16 As we can see, the results of the censored regressions are very close to the results of the
fixed effect estimates and all the previous conclusions are confirmed. The coefficients of Treatment
Multi and Treatment Choice (relative to Treatment Single) are negative and significant at the 5 percent
and at the 10 percent level, respectively. The gender-specific estimates confirm that there is no gender
difference in learning (the gender dummy is insignificant). The point estimates suggest that men
adapted better to Treatment Multi and women adapted better to Treatment Choice, but none of these
gender differences is significant.

Table 4: Impact of treatments on total points

Group-specific estimates Gender-specific estimates
FE Censored FE Censored
Treatment Multi -22.76** -24.34** -28.39** -31.09**
(10.98) (11.44) (12.98) (13.75)
Multi×Male 10.90 13.37
(21.78) (22.49)
Treatment Choice -20.97* -21.19* -14.14 -15.51
(11.33) (12.02) (16.30) (17.26)
Choice×Male -11.63 -9.12
(23.11) (24.17)
Male -3.40 -5.49
(16.39) (17.17)
Nr. of obs. 218 218 218 218
Note: Robust standard errors are shown in parentheses. Significance levels: *** p<0.01, ** p<0.05, * p<0.1.

4.2 Propensity to multitask

To examine gender differences in the propensity to multitask, we use the results of Group 3. Table 5
describes the switching behavior of men and women in Treatment Choice. As we can see, 71 percent
of the subjects do actually switch when they are allowed to and the share of switchers is exactly the
same for men and for women. So contrary to the claims of Fisher (1999), men do not focus on a
single task any more than women do. Moreover, we can reject that women switch more often than
men (one-sided t-test; p-value=0.06).
16 Note that since there were two rounds, first difference and fixed effects estimates are equivalent.

10
 Table 5: Number of switches in Treatment Choice
Men Women All
Mean 2.50 1.74 2.16
Standard deviation 2.53 1.67 2.20
Share of switchers 0.71 0.71 0.71
Number of observations 42 35 77
Note: We excluded one subject from this analysis because he misused
the ‘Switch’ button (switched multiple times within the same second).

Table 6 displays the results of two OLS regressions where the number of switches is the dependent
variable. In Column 1, we only control for performance in Round 1, while in Column 2 we include
session and task-order fixed effects. Contrary to our expectations, performance in Round 1 does not
influence switching behavior at all; this also implies that the impact of gender on switching is not
caused by performance differences. When task order and session fixed effects are also included, the
gender difference becomes significant at the 10 percent level. In sum, the results show that if there is
any gender difference, it is men switching more than women and not the other way around.

Table 6: Regression results on propensity to switch

Dependent variable: nr. of switches (1) (2)
Male 0.76 0.92*
(0.50) (0.53)
Points in Round 1 0.00066 0.00536
(0.0046) (0.0050)
Nr. of obs. 77 77
Task order and session FE no yes
Note: Robust standard errors are shown in parentheses.
Significance levels: *** p<0.01, ** p<0.05, * p<0.1

5 Discussion and conclusions

Our results confirm that work schedules are an important determinant of productivity. We find that
multitasking significantly lowers performance in cognitive tasks compared to a sequential execution.
This suggests that the costs of switching, which include recalling the rules, details and steps executed
thus far, outweigh the benefit of a ’fresh eye’.17 Note that this effect differs from the one found by
Coviello et al. (2010). In their model, every new task takes resources away from the other active tasks
which are closer to being completed, and juggling more tasks consequently increases the average
duration of task-completion. Our results show that multitasking is bad for productivity even if one is
not concerned with average duration.
17 Subjects
clearly do expect a benefit from a ’fresh eye’. Of the majority who chose to switch, many explicitly stated
’looking at the problem with a fresh eye’ as the main reason.

11
 Subjects who could choose the amount and timing of their switches freely did only marginally
better than those forced to switch at unanticipated points in time. Performance under the self-chosen
work schedule is actually significantly lower than under the exogenously imposed sequential work
schedule. This suggests that subjects fail to choose their own schedule optimally. This finding of
inability to organize one’s own work optimally is not unprecedented. For example, Ariely & Werten-
broch (2002) find that students who can set their own deadlines perform worse than those forced to
adhere to equally spaced deadlines. Another possible explanation is that even though subjects choose
the best schedule possible, their performance takes a hit due to the cognitive cost of planning. In a
sense, subjects in Treatment Choice had to perform not two but three cognitive tasks: solving the
Sudoku, solving the Word Search, and optimizing their schedule. It is difficult to distinguish between
these explanations as the number of switches is potentially endogenous to performance.18 The hy-
pothesis that additional cognitive effort is at the root of the performance impact is however supported
by the fact that the average number of switches in Treatment Choice is only 2.16, but subjects still fall
back almost as much as subjects in Treatment Multi who were forced to switch four time and could
not anticipate the timing of the switches. Whichever explanation is correct, the results are not in favor
of self-imposed work schedules.
We do not find any evidence for gender differences in the ability to multitask. Besides, the share
of switchers is exactly the same for men and women and the average number of switches is higher
for men. Thus, the results contradict the claims of Fisher (1999): if men think so much more linearly
than women, why don’t they insist more on a sequential schedule? Moreover, why is it that women
do not adapt better to multitasking than men when forced to alternate? In sum, the view that women
are better at multitasking is not supported by our findings.

References
Ariely, D. & Wertenbroch, K. (2002). Procrastination, deadlines, and performance: Self-control by
precommitment. Psychological Science, 13(3), 219–224.

Arrington, C. M. & Logan, G. D. (2004). The cost of a voluntary task switch. Psychological Science,
15(9), 610–615.

Coviello, D., Ichino, A., & Persico, N. (2010). Too many balls in the air: The impact of task juggling
on workers’ productivity. Working paper.

Criss, B. R. (2006). Gender differences in multitasking. National Undergraduate Research Clearing-

house, 9.
18 Although we find no impact of Round 1 performance, the number of switches might still be endogenous with respect
to learning or tiredness effects.

12