Biometric access-control systems—those using fingerprints, voice print

Bio access works by degrees of similarity, not identity (so say a 90% or above match will be allowed)
Even the same finger will rarely leave exactly identical prints (so 100% match will be rare)
We can adjust to minimize refusals of access to legitimate access-seekers (say 75% and above should be allowed) but such adjustments increase the likelihood of admitting impostors (now, it is more likely that an impostor will match because he has to match only 75% with someone).

The argument only talks about adjustment of reward (legitimate seeker allowed) vs risk (impostor coming in).
If you increase reward, risk increases.

(A) If a biometric access-control system were made to work by identity, it would not produce any correct admittance decisions.

We know 100% match is rare. But it is not impossible. So we cannot say that there would be no correct admittance.

(B) If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers.

Correct. If bio access system is reliable in preventing impostors (it looks for a high degree of match say 98%), it will sometimes turn away legitimate seekers because same finger will not produce the same result every time. If there is no risk, there is likely to be low reward too. So all legitimate seekers may not be allowed every time.

(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access.

We don't know what to give more weightage - risk or reward - so we cannot say when bio access is appropriate. The argument only says that bio access balances risk and reward. Whether we find it appropriate in certain cases or not, the argument doesn't say.

(D) Nonbiometric access-control systems—based, for example, on numerical codes—are less likely than biometric ones to admit impostors.

Irrelevant. No discussion on non bio access.

(E) Anyone choosing an access-control system should base the choice solely on the ratio of false refusals to false admittances.

Irrelevant. The argument doesn't say how to choose an access system (bio or non bio)

Answer (B)
_________________

I am trying to generalize all similar testing problems.

There are 4 possible responses from any such control system:

1. True positive ( someone is imposter and identifying him as imposter)
2. False positive ( someone is NOT an imposter, yet identifying him as imposter)
3. True negative (someone is NOT an imposter, and identifying him as NOT an imposter)
4. False negative (someone is an imposter, and identifying him as NOT an imposter)

1 and 3 are not the problem since the system identifies correctly, but 2 and 4 are problematic cases.

When the system is made less sensitive (i.e. the matching is accepted for lesser similarity), the system tends to allows access to both imposters and NOT imposters more easily. Hence count of 2 will decrease and count of 4 will increase. This is what is stated in the argument.

The above also means that when the system is made more sensitive (i.e. matching is accepted for more stringent similarity), the system tends to restrict access for both imposters and NOT imposters. Hence count of 2 will increase and count of 4 will decrease. This is option B.

Biometric access-control systems—those using fingerprints, voice prints, etc., to regulate admittance to restricted areas—work by degrees of similarity, not by identity. After all, even the same finger will rarely leave exactly identical prints. Such systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors.
Which of the following conclusions is most strongly supported by the information above?
(A) If a biometric access-control system were made to work by identity, it would not produce any correct admittance decisions. This is a very strong statement. Notice- "it would not produce". The premise says that they do not work by identity, but, we do not know wheather it will produce correct results or not if it works by identity.
(B) If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers. - According to the premise, if the system is adjusted to minimize legitimate refusals and increase the likelihood of admitting imposters, the converse should also be true. This is stated in this answer choice.
(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access.- The systems can be appropriate in situations where admittance of imposters can be a major problem also. One can do this by the adjusting the systems. Also, notice "only" here which signifies the strength of the statement.
(D) Nonbiometric access-control systems—based, for example, on numerical codes—are less likely than biometric ones to admit impostors. This is irrelevant, the non-biometric access is out of scope of the argument.
(E) Anyone choosing an access-control system should base the choice solely on the ratio of false refusals to false admittances. This is also irrelevant. Choosing based on the ratio of refusal to false admittance is out of scope.
IMO, it is (B). What is OA?

I second B as well.

IF the system is adjusted, something will happen. C talks about where the system can be used after adjustments. I stick to the IF and choose B. Rest of the choices are losers.

Agreed on B but one query on cause and effect!!
if x causes y
can we say when not y occurs not x is there

argument : systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors.

choice :
If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers.

im usually confused in this relation kindly shed some light on this

can we say when not y occurs not x is there
argument : systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors.
choice :
If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers.
---------------------------------------------------------------------------------------------------------------
Usually, for cause and effect, it is helpful when we dissect the argument apart:
A= Decrease in refusal to real users.
B= Increase in imposter admission.

In this case, A causes B (Given from the stem) We can not deny this.

The possible variations are:
1) Increase in imposter admission -> decrease in refusal to real users (Given)
2) Increase in refusal to real users-> decrease in imposter admission (Derived statement that must be true based on given)

The answer choice actually reiterates 2). Hence B.
For Cause and effect arguments, it is good to write a couple of words with increase and decrease arrows and understand the relationship prior to reading the answer choices. This greatly helps (atleast to me), to eliminate the incorect ones. This question might not warranty a diagram or notes, however, some of the complex cause and effect arguments (similar to puzzles) might do.
Hope this helps!!!!!!

I fail to understand the argument
can someone explain in simple words please?

chetan2u could u pls help with this one. Seems tricky

Posted from my mobile device

Biometric access-control systems—those using fingerprints, voice prints, etc., to regulate admittance to restricted areas—work by degrees of similarity, not by identity. After all, even the same finger will rarely leave exactly identical prints. Such systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors.

Which of the following conclusions is most strongly supported by the information above?


(A) If a biometric access-control system were made to work by identity, it would not produce any correct admittance decisions.

(B) If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers.

(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access.

(D) Nonbiometric access-control systems—based, for example, on numerical codes—are less likely than biometric ones to admit impostors.

(E) Anyone choosing an access-control system should base the choice solely on the ratio of false refusals to false admittances.


The passage talks about certain biometric systems that use finger print, voice etc to provide access. its uses similarity instead of identity. This minimizes refusals of access to legitimate users but also increases the likelihood of admitting impostors.
question asks us to draw an inference.

a) passage doesnt even talk about how identity based systems work, so we cannot infer this .
b) this is the coorect answer. 'Sometimes' is the key word here and tells us that there is a possibility that it may turn away legitimate users as stated in passage.
c) even though this looks tempting, we cannot infer this. In inference questions it very important to stick to the info provided in the passage. For eg what about situations in which admittance of imposters and denial of legitimate users are equally likely. Hence cannot infer this
d)Out of scope totally.
e)The choice isnt even mentioned in the passage. It just talks about biometric access systems.

My doubt is, if the system will reliably prevent impostors from being admitted [I assume, it means ALL the time], to achieve this, we would have to use identity based approach rather than similarity based. This is so because, identity based approach is the ONLY way to make sure that an imposter is NEVER allowed in the restrictive area. But then, as a consequence, Genuine people will MOST OF THE TIMES [mostly] rejected, as the prints are same RARELY.

So, I would have accepted the options if it said, mostly insead of sometimes. And that's the reason I rejected this option.

Can you please help.

Biometric access-control systems—those using fingerprints, voice prints, etc., to regulate admittance to restricted areas—work by degrees of similarity, not by identity. After all, even the same finger will rarely leave exactly identical prints. Such systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors.

Which of the following conclusions is most strongly supported by the information above?

(A) If a biometric access-control system were made to work by identity, it would not produce any correct admittance decisions.

No info. All we know from the passage are details about similarity. We should not take occasional rejections of legitimate access-seekers (on the basis of SIMILARITY) to mean that NOBODY would ever gain access based on IDENTITY. That's a large and unjustified leap in logic.

(B) If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers.
Correct.

(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access.

(D) Nonbiometric access-control systems—based, for example, on numerical codes—are less likely than biometric ones to admit impostors.

We don't know anything about non-biometric systems. Could be true, could not be.

(E) Anyone choosing an access-control system should base the choice solely on the ratio of false refusals to false admittances.

No. Possible sure? Hardly even plausible even if we were to think about this in reality. Lots of factors affect choice of such a (presumably expensive system). <--- See what I did there? Cost is another factor.
_________________

gmatter923, perhaps this is too little too late, but I can try to help. The problem lies in your own assumption. It does not say that the system will prevent imposters all the time. We don't know if the success rate is 100%.

According to Merriam-Webster,
Definition of some (Entry 1 of 6)
1 : being an unknown, undetermined, or unspecified unit or thing
2a : being one, a part, or an unspecified number of something (such as a class or group) named or implied
2b : being of an unspecified amount or number

The answer means that the person has shopped online at least once. That's it. The answer doesn't mean that the person shops online a small number of times. The person could very well shop online most of the times.

If you can infer that something happens most of the times, it certainly happens at least once.

Say, x = 90.
So, is x >50? Yes.
Is x >= 1? Also yes.

Hi Experts, ]AndrewN KarishmaB HaileyCusimano

Please can you help me with the below query.

The last two sentences of the passage states that "Such systems can be adjusted to minimize refusals of access to legitimate access-seekers. Such adjustments, however, increase the likelihood of admitting impostors." Thus why would Option C be incorrect when we consider these two statements.

I was stuck between B & C and ended up choosing C because of the above.

Here's (C).

(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access.

In order for what choice (C) says to be true, it would have to be true that there is an equal or close to equal balance between solving the problem of mistaken refusal of access and creating the problem of admittance of imposters such that once the mistaken refusal of access problem is solved to a reasonable degree, there is a significant probability that imposters will gain admittance.

So, notice that the passage says only that the adjustments "increase the likelihood of admitting impostors." It's doesn't say that admittance of imposters is significantly increased when refusal of access is largely solved.

Thus, for all we know from what the passage says, refusal of access can be almost 100 percent solved with the possibility of admittance of imposters increased only negligibly, meaning that the problem of admittance of imposters barely exists when the problem of refusal of access also barely exists.

In such a case, what (C) says would not follow, since even if admittance of imposters is much more of a problem than mistaken refusal of access, the tradeoff involved in making the adjustments necessary for making the use of a biometric access-control system practical would be worthwhile given the low probability that an imposter will gain admittance even after the adjustments have been made.
_________________

Hello, KittyDoodles. It looks like Marty beat me again, but I will reproduce my response anyway. For reference, here is answer choice (C):


Note the extreme condition imposed on the use of biometric access-control systems: only in a specific type of situation. The passage does not allow us to reasonably infer that these systems should not be used in other situations. The last two lines of the passage should not be interpreted as saying that illegitimate access is granted as frequently as legitimate access. It could be the case that by relaxing the checks somewhat for degrees of similarity, the likelihood of someone gaining illegitimate access might go from virtually zero to, say, one percent. Meanwhile, those seeking legitimate access could gain admittance in perhaps ninety-nine percent of cases, a pretty fair compromise (in terms of convenience) for slightly reduced security.

Perhaps the answer choice makes more sense now. Thank you for thinking to ask.

- Andrew
_________________

P: Biometric access systems work by degrees of similarity, not by identity.
P: This is because: even the same finger will rarely leave exactly identical prints.
P: Using system setting tweaks - the access system can be adjusted to minimise refusals of legitimate access-seekers (people who are authorised to enter/ access work).
P: But such adjustments increase the likelihood of admitting impostors.

Which of the following conclusions is most strongly supported by the information above?

(A) If a biometric access-control system were made to work by identity, it would not produce any correct admittance decisions.: would not produce any correct admittance decisions. this is a bit extreme as it uses the words of the passage but twists it in such a way that it goes too far.

The passage is saying: it works on similarity not identify because same prints are not left by same finger. further we dont know what would happen if a system worked on identity per say. this is because we do not have any info on it.

(B) If a biometric access-control system reliably prevents impostors from being admitted, it will sometimes turn away legitimate access-seekers: because it uses similarity it can sometimes reject authorised personnel. This is what the stimulus is saying.

as mentioned already - According to the premise, if the system is adjusted to minimise legitimate refusals and increase the likelihood of admitting imposters, the converse should also be true. This is stated in this answer choice.

(C) Biometric access-control systems are appropriate only in situations in which admittance of impostors is less of a problem than is mistaken refusal of access: takes it too far. its saying that we should only use the access systems when admitting an imposter is less of an issue than an authorised individual being rejected. The passage doesn't say this. further, you can adjust the settings.

(D) Nonbiometric access-control systems—based, for example, on numerical codes—are less likely than biometric ones to admit impostors.: outside info.

(E) Anyone choosing an access-control system should base the choice solely on the ratio of false refusals to false admittances: outside info, not mentioned previously.

I hope my problem-solving approach corroborates the selection of B as the correct choice. So here is my two cents.

Consider the two events
A: Decrease in rejection to actual users [reward] &
B: Increase in falsified entry [risk]

From the stem, we know that A caused B. This causality would mean that A is a sufficient condition for B.

∴ A(sufficient) → B(necessary)

Any paraphrased statement or "contrapositive" statement of the above version is logically valid. [Refer to "Conditional Reasoning" topic from Powerscore CR Bible.]

Hence, Not B → Not A is also valid. (This is the contrapositive form of A → B.)

In our case,
Not A: Increase in rejection of actual users
Not B: Decrease in falsified entry

This form is reflected in answer choice B.

Additional Note:
B → A (called the "false Reversal") &
Not A→ Not B (called the "false negation") are logically invalid forms.