Limits...
Assessing uncertainty in sighting records: an example of the Barbary lion.

Lee TE, Black SA, Fellous A, Yamaguchi N, Angelici FM, Al Hikmani H, Reed JM, Elphick CS, Roberts DL - PeerJ (2015)

Bottom Line: We find that asking experts to provide scores for these three aspects resulted in each sighting being considered more individually, meaning that this new questioning method provides very different estimated probabilities that a sighting is valid, which greatly affects the outcome from an extinction model.We consider linear opinion pooling and logarithm opinion pooling to combine the three scores, and also to combine opinions on each sighting.We find the two methods produce similar outcomes, allowing the user to focus on chosen features of each method, such as satisfying the marginalisation property or being externally Bayesian.

View Article: PubMed Central - HTML - PubMed

Affiliation: Mathematical Institute, University of Oxford , UK.

ABSTRACT
As species become rare and approach extinction, purported sightings can be controversial, especially when scarce management resources are at stake. We consider the probability that each individual sighting of a series is valid. Obtaining these probabilities requires a strict framework to ensure that they are as accurately representative as possible. We used a process, which has proven to provide accurate estimates from a group of experts, to obtain probabilities for the validation of 32 sightings of the Barbary lion. We consider the scenario where experts are simply asked whether a sighting was valid, as well as asking them to score the sighting based on distinguishablity, observer competence, and verifiability. We find that asking experts to provide scores for these three aspects resulted in each sighting being considered more individually, meaning that this new questioning method provides very different estimated probabilities that a sighting is valid, which greatly affects the outcome from an extinction model. We consider linear opinion pooling and logarithm opinion pooling to combine the three scores, and also to combine opinions on each sighting. We find the two methods produce similar outcomes, allowing the user to focus on chosen features of each method, such as satisfying the marginalisation property or being externally Bayesian.

No MeSH data available.


The distribution of ‘best’ estimates over 160 (5 experts scoring 32 sightings) responses, together with the 25th, 50th, and 75th percentiles.The dotted line indicates the 50th percentile (the median) and the shaded error indicates the interquartile range (the range between the 25th and 75th percentile). The 25th, 50th and 75th percentile values are provided under each plot.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4562256&req=5

fig-1: The distribution of ‘best’ estimates over 160 (5 experts scoring 32 sightings) responses, together with the 25th, 50th, and 75th percentiles.The dotted line indicates the 50th percentile (the median) and the shaded error indicates the interquartile range (the range between the 25th and 75th percentile). The 25th, 50th and 75th percentile values are provided under each plot.

Mentions: We first considered the distribution of the raw data; that is, 160 (5 experts each judging 32 sightings) responses for each sighting (see Supplemental Information 3). When simply asked whether the sighting was correct (Q1), the responses follow a nearly identical distribution to responses on whether the sighting was distinguishable (Q2), see Figs. 1A and 1B. For both Q1 and Q2, to one decimal place, half the responses lie within the conservative range of 0.7 and 0.9, centred evenly around the median of approximately 0.8. Arguably distinguishability may not vary much, but the small interquartile range for Q1 raises questions about whether it is a true representation of the diverse sighting quality (see Supplemental Information 1). The broad nature of Q1 may make it more susceptible to behavioural aspects, such as question fatigue, than specific questions such as Q2, Q3 and Q4.


Assessing uncertainty in sighting records: an example of the Barbary lion.

Lee TE, Black SA, Fellous A, Yamaguchi N, Angelici FM, Al Hikmani H, Reed JM, Elphick CS, Roberts DL - PeerJ (2015)

The distribution of ‘best’ estimates over 160 (5 experts scoring 32 sightings) responses, together with the 25th, 50th, and 75th percentiles.The dotted line indicates the 50th percentile (the median) and the shaded error indicates the interquartile range (the range between the 25th and 75th percentile). The 25th, 50th and 75th percentile values are provided under each plot.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4562256&req=5

fig-1: The distribution of ‘best’ estimates over 160 (5 experts scoring 32 sightings) responses, together with the 25th, 50th, and 75th percentiles.The dotted line indicates the 50th percentile (the median) and the shaded error indicates the interquartile range (the range between the 25th and 75th percentile). The 25th, 50th and 75th percentile values are provided under each plot.
Mentions: We first considered the distribution of the raw data; that is, 160 (5 experts each judging 32 sightings) responses for each sighting (see Supplemental Information 3). When simply asked whether the sighting was correct (Q1), the responses follow a nearly identical distribution to responses on whether the sighting was distinguishable (Q2), see Figs. 1A and 1B. For both Q1 and Q2, to one decimal place, half the responses lie within the conservative range of 0.7 and 0.9, centred evenly around the median of approximately 0.8. Arguably distinguishability may not vary much, but the small interquartile range for Q1 raises questions about whether it is a true representation of the diverse sighting quality (see Supplemental Information 1). The broad nature of Q1 may make it more susceptible to behavioural aspects, such as question fatigue, than specific questions such as Q2, Q3 and Q4.

Bottom Line: We find that asking experts to provide scores for these three aspects resulted in each sighting being considered more individually, meaning that this new questioning method provides very different estimated probabilities that a sighting is valid, which greatly affects the outcome from an extinction model.We consider linear opinion pooling and logarithm opinion pooling to combine the three scores, and also to combine opinions on each sighting.We find the two methods produce similar outcomes, allowing the user to focus on chosen features of each method, such as satisfying the marginalisation property or being externally Bayesian.

View Article: PubMed Central - HTML - PubMed

Affiliation: Mathematical Institute, University of Oxford , UK.

ABSTRACT
As species become rare and approach extinction, purported sightings can be controversial, especially when scarce management resources are at stake. We consider the probability that each individual sighting of a series is valid. Obtaining these probabilities requires a strict framework to ensure that they are as accurately representative as possible. We used a process, which has proven to provide accurate estimates from a group of experts, to obtain probabilities for the validation of 32 sightings of the Barbary lion. We consider the scenario where experts are simply asked whether a sighting was valid, as well as asking them to score the sighting based on distinguishablity, observer competence, and verifiability. We find that asking experts to provide scores for these three aspects resulted in each sighting being considered more individually, meaning that this new questioning method provides very different estimated probabilities that a sighting is valid, which greatly affects the outcome from an extinction model. We consider linear opinion pooling and logarithm opinion pooling to combine the three scores, and also to combine opinions on each sighting. We find the two methods produce similar outcomes, allowing the user to focus on chosen features of each method, such as satisfying the marginalisation property or being externally Bayesian.

No MeSH data available.