janssen et al. - escop xiv1 dating events: solving the jigsaw puzzle steve m. j. janssen, antonio g....
TRANSCRIPT
Janssen et al. - ESCoP XIV 1
Dating Events: Solving The Jigsaw Puzzle
Steve M. J. Janssen,Antonio G. Chessa, and Jaap M. J. Murre
University of Amsterdam
Janssen et al. - ESCoP XIV 2
Jigsaw puzzle
Janssen et al. - ESCoP XIV 3
Pieces of a puzzle
• Pieces with a lot of details • Exact date• Pieces with a lot of details• Filler pieces
• Exact date• Context
• Pieces with a lot of details• Filler pieces• Connected pieces
• Exact date• Context• Related events
• Pieces with a lot of details• Filler pieces• Connected pieces • Corner and edge pieces
• Exact date• Context• Related events• Finer time scales
Janssen et al. - ESCoP XIV 4
Memory for time theories
• Distance is the elapsed time (i.e., relative time)• Distance is the elapsed time (i.e., relative time)• Location is the placement of an event in time (i.e.,
absolute time)
• Distance is the elapsed time (i.e., relative time)• Location is the placement of an event in time (i.e.,
absolute time)• Serial order is the reference to fact that an event
occurred before of after another event
Janssen et al. - ESCoP XIV 5
Integrated Model Of Memory For Time (Janssen et al., in press, Mem & Cog)
• Primary temporal information, such as day of month. When there is direct temporal information of an event available, then people can determine the exact date that the event occurred. However, this direct information is forgotten easily.
• Secondary temporal information, such as context. When there is indirect temporal information available, then people prefer to date the event in the absolute time format. This indirect information is forgotten less easily.
• When there is neither primary or secondary temporal information available, then people prefer to date the event in the relative time format.
Janssen et al. - ESCoP XIV 6
Integrated Model Of Memory For Time
Type of information Format preference Availability
Primary Information: Who, What, Where, and When (time of day, day of week, etc)
Secondary Information: Context and Related events
Is there primary temporal information?
Is there distinctive secondary temporal information?
How familiar is the event? How many intervening events?
Absolute time format
Absolute time format
Relative time format
No
No
Yes
Yes
Age Event
Very recent
Recent
Remote
Janssen et al. - ESCoP XIV 7
Experiment and Hypothesis
• Do people use different temporal information to date different events?
• In this experiment, we will not look at format preference, but at different time scales
• Do people use different temporal information to date different events?
• Do people recall for some events the year that event occurred better, while for other events the day of the week?
• Do people use different temporal information to date different events?• Do people recall for some events the year that event occurred better,
while for other events the day of the week?• Are the time scales independent for each event?
8Janssen et al. - ESCoP XIV
Participants
• http://memory.uva.nl/• http://memory.uva.nl/• April 2004 – March 2005
• http://memory.uva.nl/• April 2004 – March 2005• N = 3974
• http://memory.uva.nl/• April 2004 – March 2005• N = 3974• M age = 42.23 yrs• Majority female (57.2%)
• http://memory.uva.nl/• April 2004 – March 2005• N = 3974• M age = 42.23 yrs• Majority female (57.2%)• Majority equivalent of university or college degree
(58.9%)
• http://memory.uva.nl/• April 2004 – March 2005• N = 3974• M age = 42.23 yrs
9
MaterialsTsunami Sunday - December 26, 2004 - 02:59
Bernhard Wednesday - December 1, 2004 - 18:50
Van Gogh Tuesday - November 2, 2004 - 9:00
Juliana Saturday - March 20, 2004 - 5:50
Madrid Thursday - March 11, 2004 - 07:39
Amalia Sunday - December 7, 2003 - 17:01
Claus Sunday - October 6, 2002 - 19:00
Fortuyn Monday - May 6, 2002 - 18:06
WA&M Saturday - February 2, 2002 - 10:30
WTC Tuesday - September 11, 2001 - 14:46
Brood Wednesday - July 11, 2001 - 13:30
Enschede Saturday - May 13, 2000 - 15:34
Bijlmer Sunday - October 4, 1992 - 18:35
Janssen et al. - ESCoP XIV 10
Procedure
• Design from Friedman & Wilkins (1985)• Design from Friedman & Wilkins (1985)• Semi-random selection of five events
• Design from Friedman & Wilkins (1985)• Semi-random selection of five events• For each event, participants had to date the event on
only one scale
• Design from Friedman & Wilkins (1985)• Semi-random selection of five events• For each event, participants had to date the event on
only one scale• Five scales: Year (1989-2004/2005), Month (Jan – Dec),
Day Of Month (1-31), Day Of Week (Mon – Sun), and Hour (0:00 – 24:00)
• Design from Friedman & Wilkins (1985)• Semi-random selection of five events• For each event, participants had to date the event on
only one scale• Five scales: Year (1989-2004/2005), Month (Jan – Dec),
Day Of Month (1-31), Day Of Week (Mon – Sun), and Hour (0:00 – 24:00)
• For example: “In which month was Princess Amalia born?”
Janssen et al. - ESCoP XIV 11
Proportion correct
)1
)1((1
)1(n
CPn
nRP
P(C=1) is the proportion correct
P(R=1) is the adjusted proportion correct
n is the number of options per scale
12
Proportion correct
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Janssen et al. - ESCoP XIV 13
Proportion correct
• Friedman’s Q(4)=17.60, p=.01, Kendall’s W=.440 -> Dependent
• Friedman’s Q(4)=17.60, p=.01, Kendall’s W=.440 -> Dependent
• Without year and DOM scale: Friedman’s Q(2)=5.60, p>.05, Kendall’s W=.280 -> Independent
• Friedman’s Q(4)=17.60, p=.01, Kendall’s W=.440 -> Dependent
• Without year and DOM scale: Friedman’s Q(2)=5.60, p>.05, Kendall’s W=.280 -> Independent
• Only remote events: Friedman’s Q(4)=6.53, p>.05, Kendall’s W=.272 -> Independent
Janssen et al. - ESCoP XIV 14
Distance
21
21
),(1
),( 1
n
ACdN
ACd
N
ii
‘d(C,A) is the adjusted distance
d(C,Ai) is the absolute distance per answer
N is the number of answers
n is the number of options per scale
15
Distance
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Year Month DOM DOW Hour
Tsunami
Bernhard
Van Gogh
Juliana
Madrid
Amalia
Claus
Fortuyn
WA&M
WTC
Brood
Enschede
Bijlmer
Janssen et al. - ESCoP XIV 16
Distance
• Friedman’s Q(4)=17.60, p=.01, Kendall’s W=.440 -> Dependent
• Friedman’s Q(4)=17.60, p=.01, Kendall’s W=.440 -> Dependent
• Without month scale: Friedman’s Q(3)=5.40, p>.05, Kendall’s W=.180 -> Independent
Janssen et al. - ESCoP XIV 17
Discussion
• We found that the time scales were independent, when we only looked at events that occurred more than two years ago
• Or when we omitted one of the time scales
Janssen et al. - ESCoP XIV 18
Incomplete puzzle
For some events, people recall the exact date, while for other events only the period in which they occurred. For other events, people remember that they happened just after another event, but most events are forgotten. In contrast to a real jigsaw puzzle, the puzzle of event dating can never be completely solved
Janssen et al. - ESCoP XIV 19
Thank you for your attention
• Steve Janssen• http://memory.uva.nl/• [email protected]