WHEN A HAYSTACK CAN HAVE MANY NEEDLES
Techniques of Lead-pool Analysis and Hit-Rate Enhancement in Disappearance Cases
By Greg Hooper, CFLI
Hooper Investigative Associates, Santa Rosa, CA.
Email: pihoop@mindspring.com

 

The purpose of this article is to find out if other investigators have any experience with hotline lead-pool analysis or tip-line hit-rate enhancement techniques comparable to that presented below, and to hopefully receive their (your) commentary. Below is an abbreviated version of an article in the next (Sept/Oct) issue of PI Magazine.

How can an entire "haystack" of tips and leads in a disappearance case become useful as opposed to just the proverbial needle hopefully hidden within it? With further testing of techniques as suggested below, the 'haystack' itself may become helpful in finding not just one, but several 'needles' of info that would close a disappearance case.

First to be mentioned though: posting a civilian hotline number or family phone number along with the proper law enforcement number on the TRAK flyers is important because many types of possible informants will not call a police line. These can include an abductor's associates who don't like what's happening to a victim (prostitution, other harm) or who may be troubled by the wrong that was done, but who fear being traced or ID'd somehow by the police. The fear can be because such potential informants may be criminals themselves or may fear retribution from the suspect if s/he (the informant) is ID'd or traced and the suspect somehow finds out about it. Such a caller may feel more in control by calling civilians.

Finally, suspects themselves are known to call to engage in predatory interaction with the friends or family, or try to give misleading reports or suggestions (this kind of call can actually be most helpful). They are not as likely to call a law enforcement number as a civilian line.

For friends and family who are not up to receiving such calls, they can publish a separate phone number to an answering machine. Then the recorded calls can be available for private sector professional analysis as well as to law enforcement. Or an investigator or a trained friend can receive and log (and possibly record, depending on local laws) such calls.

But now to the nitty-gritty of the matter for me; I've only had very limited opportunities to try this in a systematic way since my last disappearance case. So it is still largely theoretical at this point.
This approach assumes that the person who answers the family phone or civilian hotline is taking notes on lead-sheets according to standard disappearance hotline protocol. (It helps greatly if they can record too ­ check with local authorities.)

After receiving a tip, advice, suggestion or whatever, each caller's statement can be rated in three or four categories, the something like following. We'll see how these may A- reveal something important about the caller and B- speed up prioritizing of the leads.

 

1- Identifying Information: How much identifying information did the caller give about him/herself (strong, weak, none, or apparently deceptive)? That is, how much identifying info did the caller give so that you can actually look them up (the caller that is), and call and or find them?

2- Descriptive Accuracy: How descriptively accurate is the caller's info about the victim? Is the caller's descriptive accuracy of the victim in their sighting/tip strong or weak? Is the caller accurately describing the missing person or possibly mistaking someone else for the victim?

3- Case-Familiarity: Does the caller have strong or weak case-knowledge or familiarity with the family or people involved, especially with the victim?

4- Case-related Logic: How logical is the caller's information or suggestions with regard to what is and is not likely about the victim. That is, considering the victim and his or her last known circumstances, is the caller giving you a sighting/tip/suggestion that is logical or likely? Is the caller saying that s/he saw the victim in Antarctica with the British polar bear ice swimming society? How likely is that tip if the victim is a ten year old who can't stand the cold? That would be weak case-logic, not very likely. What if the caller said the victim was seen climbing a tree in a neighbor's back forty. That might be quite possible and could be strong case-logic.



If the caller is giving a suggestion, not a sighting/tip (and this goes for psychics), how likely or logical would the advice be if the caller suggests looking in Antarctica? If on the other hand, the caller suggests looking in the tree houses of neighbors and the child is known to appreciate tree houses, that could be strong case-logic.

Now, regardless of the type of specific scoring system you use ("strong vs. weak" or a ten point system, whatever you want to try), the way these four scores in each lead relate (their ratios) to each other can be incredibly meaningful about the person calling.

What if you just reviewed or received a called-in lead that was high in descriptive accuracy (they accurately described the victim), and case logic (there information seemed logical) and at least mid range on ID (they gave a full name or a call-back phone number) but low case familiarity (they didn't speak like they knew much about the victim or the case)? With those scores, this lead goes to the top of your heap for first follow-up efforts because you've got good sounding info from someone who doesn't otherwise know the victim or the case. Because of such low case familiarity, that caller is less likely to be concocting any of the accurate details and strong logical possibilities.

What if you got a lead with high case-familiarity (they spoke like they knew a lot about the victim or the case) and gave strong descriptive accuracy (they accurately described the missing person), but they gave low ID (only a first name, no call-back number) and low case-logic (the location they say they saw the victim or where they say to look is not very logical or likely to be true)? This one also goes to the top of the pile for it's suspicious nature. It could be a case-related suspect trying to muddy up his trail. The exact time of the call should be noted to facilitate a trap and trace as soon as possible. And hopefully it will have been recorded. Often a lie or misleading information as well as voice inflection and innuendo can reveal the information hidden.

What if you got a call from a 'curious member of the public' who just wanted to know how the family was holding up, and how they were getting on, and the kinds of things they are saying, and how they are getting by and what they are doing to cope? And what if this caller acted like he didn't know much about the case or victim but actually sounded (said little things) like they knew more, showing higher case-familiarity than they would want you to think? And what if they gave weak or no ID? Strong case-familiarity and weak ID by themselves are fairly good combination for your suspicious stack.

What if the caller's statement showed high ID but low descriptive accuracy, low case-familiarity and low case-logic (the only definite info you are getting is who they are)? It is probably somebody who doesn't get enough attention or doesn't get out much, and you can bury it.

This type of objective scoring will also free you from the need to be judgmental about psychic suggestion/leads. All you have to do is apply these or similar objectifying criteria. That is, whether a call-in lead is psychic or not, if the statement has high descriptive accuracy, case logic, ID and low case-familiarity (which is actually the best combination), why throw it out? Throw it to the top of the heap with the other good ones and hit the road with it.

Finally, such a scoring system is not truly objective of course. The scores on any given lead will depend a lot on the individual taking the call or receiving the info. And, many lead/tips are not going to be easy to apply these four (or similar) measures to (one example below). Fortunately, such a scoring system does not require a great deal of accuracy or uniformity to be helpful in isolating the leads with high potential from the ones that just seem strong or suspect on the surface.

This article was written particularly because of work experience in stranger abductions. However, here is an example (one of only a few I have since formulating this approach) from a current parental abduction. It involves a tip received at a civilian hotline center recently from an anonymous e-mailer:

"Do you have to tell the person who has the child where the tip came from? I'm just worried if [abductors name] finds out who told you about him. Do the police tell the person where the tip came from? Have to be concerned with my own safety. But I'm worried about [child's name] too. [abductors name] dyed his hair blond and tells people his name is [his new a.k.a.]. Also, what if the police just go pick him up at the place I tell you about, which only I would have known. Then he'll know who told them about it. Do they [the police] care about that? They are in the [city name] area but I have to think about this" [about giving the needed details].

First of all, its easy to see from his (the e-mailer's user ID implies that he is a male) nervousness and questions that we probably wouldn't have heard from him if the only number he knew to contact in the case was the law enforcement agency involved. For someone like this (who may also be criminal), he feels much more in control and safer talking to civilians.

So, how does this tip score on the above four suggested criteria? In descriptive accuracy, he only refers to the girl by name once and with no description. But the e-mailer's nervousness indicates that he is not just calling to get involved because he's bored and he just might have seen the right person. He "sounds" like he's talking about our missing subjects, considering other details. We see from things implied that he has personal familiarity with the subject. We too happen to know that the subject isn't the kind of person that you would want to rat on and that the child is not safe with him either. We also know from a recent DMV photo in another state that he did cut and dye is hair blond.

So I myself would give this a "high or strong" score in descriptive accuracy even without a lot of description. What about the tipster's ID? That's obviously "low or weak" in this case, though at least "not deceptive" sounding. But the caller has a good sounding reason, other than intrigue or meddling, for anonymity.

How logical does this information sound (case-related logic)? It is "high" to me, even without much detail in the message. This is because we know that the subject has a history and many contacts in the city area mentioned.

And for me, case familiarity in this message is "medium to low" because the only familiarity he indicated having about the case was based on his personal familiarity with the subject (implied from things said about him), not from knowing the family or news accounts of the abduction where the child came from. Again, this would indicate that the caller (e-mailer) is not concocting anything from details already known to him as would a suspect or bored trouble-making lamer.

So, with high descriptive accuracy, high case-related logic, medium to low case familiarity we have (in my opinion) a strong lead, even with low or weak ID. Strong ID would have been even better. (The fact is, I was only recently consulted on this case and only given this and four other recent, equally intense sounding, but rather contrary leads. I favored this one because of the above scoring. I've since received confirmation that it was the right choice to follow. (The primary investigator identified the neighborhood of the e-mailer and other witnesses. For the reporting party's (the e-mailer's) safety, when we make the recovery, we'll do it as a 911 call by an on-site 'witness' with a TRAK flyer in hand. We will arrange it so that the 'the witness' making the call will also be known to the abductor as the reporting party so that he will not make a connection to the actual reporting party. Only we will know who that is.

This writer is most interested to hear from other investigators about this or similar approaches, if any, to lead-pool hit-rate enhancement.

Greg Hooper, CFLI, PhD(abd)
Hooper Investigative Associates, Santa Rosa, CA.
Email: pihoop@mindspring.com
Web Site: http://www.threat-assessment.net
Phone: 707-570-0818
Member- National Association of Investigative Specialists, Association of Threat Assessment Professionals, registered volunteer of HomeFires.org ­ volunteer PIs for missing children.

Copyright: 2001, Greg Hooper.
All rights reserved.