“It [ACL] specifies a list of users [subjects] who are allowed access to each object” CBK, p. 188
A capability table is incorrect. “Capability tables are used to track, manage and apply controls based on the object and rights, or capabilities of a subject. For example, a table identifies the object, specifies access rights allowed for a subject, and permits access based on the user’s posession of a capability (or ticket) for the object.” CBK, pp. 191-192. The distinction that makes this an incorrect choice is that access is based on posession of a capability by the subject.
To put it another way, as noted in AIO3 on p. 169, “A capabiltiy table is different from an ACL because the subject is bound to the capability table, whereas the object is bound to the ACL.”
An access control matrix is incorrect. The access control matrix is a way of describing the rules for an access control strategy. The matrix lists the users, groups and roles down the left side and the resources and functions across the top. The cells of the matrix can either indicate that access is allowed or indicate the type of access. CBK pp 317 -318.
AIO3, p. 169 describes it as a table if subjects and objects specifying the access rights a certain subject possesses pertaining to specific objects.
In either case, the matrix is a way of analyzing the access control needed by a population of subjects to a population of objects. This access control can be applied using rules, ACL’s, capability tables, etc.
A role-based matrix is incorrect. Again, a matrix of roles vs objects could be used as a tool for thinking about the access control to be applied to a set of objects. The results of the analysis could then be implemented using RBAC.
References:
CBK, Domain 2: Access Control. AIO3, Chapter 4: Access Control

The underlying problem for a company with a lot of turnover is assuring that new employees are assigned the correct access permissions and that those permissions are removed when they leave the company.
Selecting the best answer requires one to think about the access control options in the context of a company with a lot of flux in the employee population. RBAC simplifies the task of assigning permissions because the permissions are assigned to roles which do not change based on who belongs to them. As employees join the company, it is simply a matter of assigning them to the appropriate roles and their permissions derive from their assigned role. They will implicitely inherit the permissions of the role or roles they have been assigned to. When they leave the company or change jobs, their role assignment is revoked/ changed appropriately.
Mandatory access control is incorrect. While controlling access based on the clearence level of employees and the sensitivity of obects is a better choice than some of the other incorrect answers, it is not the best choice when RBAC is an option and you are looking for the best solution for a high number of employees constantly leaving or joining the company.
Lattice-based access control is incorrect. The lattice is really a mathematical concept that is used in formally modeling information flow (Bell-Lapadula, Biba, etc). In the context of the question, an abstract model of information flow is not an appropriate choice. CBK, pp. 324-325.
Discretionary access control is incorrect. When an employee joins or leaves the company, the object owner must grant or revoke access for that employee on all the objects they own. Problems would also arise when the owner of an object leaves the company. The complexity of assuring that the permissions are added and removed correctly makes this the least desirable solution in this situation.
References
Alll in One, third edition page 165 RBAC is discussed on pp. 189 through 191 of the ISC(2) guide.

RBAC is sometimes also called non-discretionary access control (NDAC) (as Ferraiolo says “to distinguish it from the policy-based specifics of MAC”). Another model that fits within the NDAC category is Rule-Based Access Control (RuBAC or RBAC).
Most of the CISSP books use the same acronym for both models but NIST tend to use a lowercase “u” in between R and B to differentiate the two models.
You can certainly mimic MAC using RBAC but true MAC makes use of Labels which contains the sensitivity of the objects and the categories they belong to. No labels means MAC is not being used.
One of the most fundamental data access control decisions an organization must make is the amount of control it will give system and data owners to specify the level of access users of that data will have. In every organization there is a balancing point between the access controls enforced by organization and system policy and the ability for information owners to determine who can have access based on specific business requirements. The process of translating that balance into a workable access control model can be defined by three general access frameworks:
Discretionary access control Mandatory access control Nondiscretionary access control
A role-based access control (RBAC) model bases the access control authorizations on the roles (or functions) that the user is assigned within an organization. The determination of what roles have access to a resource can be governed by the owner of the data, as with DACs, or applied based on policy, as with MACs.
Access control decisions are based on job function, previously defined and governed by policy, and each role (job function) will have its own access capabilities. Objects associated with a role will inherit privileges assigned to that role. This is also true for groups of users, allowing administrators to simplify access control strategies by assigning users to groups and groups to roles. There are several approaches to RBAC. As with many system controls, there are variations on how they can be applied within a computer system.
There are four basic RBAC architectures: 1. Non-RBAC: Non-RBAC is simply a user-granted access to data or an application by traditional mapping, such as with ACLs. There are no formal “roles” associated with the mappings, other than any identified by the particular user.
2. Limited RBAC: Limited RBAC is achieved when users are mapped to roles within a single application rather than through an organization-wide role structure. Users in a limited RBAC system are also able to access non-RBAC-based applications or data. For example, a user may be assigned to multiple roles within several applications and, in addition, have direct access to another application or system independent of his or her assigned role. The key attribute of limited RBAC is that the role for that user is defined within an application and not necessarily based on the user’s organizational job function.
3. Hybrid RBAC: Hybrid RBAC introduces the use of a role that is applied to multiple applications or systems based on a user’s specific role within the organization. That role is then applied to applications or systems that subscribe to the organization’s role-based model. However, as the term “hybrid” suggests, there are instances where the subject may also be assigned to roles defined solely within specific applications, complimenting (or, perhaps, contradicting) the larger, more encompassing organizational role used by other systems. 4. Full RBAC: Full RBAC systems are controlled by roles defined by the organization’s policy and access control infrastructure and then applied to applications and systems across the enterprise. The applications, systems, and associated data apply permissions based on that enterprise definition, and not one defined by a specific application or system. Be careful not to try to make MAC and DAC opposites of each other –they are two different access control strategies with RBAC being a third strategy that was defined later to address some of the limitations of MAC and DAC.
The other answers are not correct because:
Mandatory access control is incorrect because though it is by definition not discretionary, it is not called “non-discretionary access control.” MAC makes use of label to indicate the sensitivity of the object and it also makes use of categories to implement the need to know.
Label-based access control is incorrect because this is not a name for a type of access control but simply a bogus detractor. Lattice based access control is not adequate either. A lattice is a series of levels and a subject will be granted an upper and lower bound within the series of levels. These levels could be sensitivity levels or they could be confidentiality levels or they could be integrity levels.
Reference(s) used for this question:
All in One, third edition, page 165. Ferraiolo, D., Kuhn, D. & Chandramouli, R. (2003). Role-Based Access Control, p. 18.
Ferraiolo, D., Kuhn, D. (1992). Role-Based Access Controls. http://csrc.nist.gov/rbac/Role_Based_Access_Control-1992.html
Schneiter, Andrew (2013-04-15). Official (ISC)2 Guide to the CISSP CBK, Third Edition : Access Control ((ISC)2 Press) (Kindle Locations 1557-1584). Auerbach Publications. Kindle Edition. Schneiter, Andrew (2013-04-15). Official (ISC)2 Guide to the CISSP CBK, Third Edition : Access Control ((ISC)2 Press) (Kindle Locations 1474-1477). Auerbach Publications. Kindle Edition.

Access decisions are made based on the clearance of the subject and the sensitivity label of the object.
Example: Eve has a “Secret” security clearance and is able to access the “Mugwump Missile Design Profile” because its sensitivity label is “Secret.” She is denied access to the “Presidential Toilet Tissue Formula” because its sensitivity label is “Top Secret.”
The other answers are not correct because:
Discretionary Access Control is incorrect because in DAC access to data is determined by the data owner. For example, Joe owns the “Secret Chili Recipe” and grants read access to Charles.
Role Based Access Control is incorrect because in RBAC access decsions are made based on the role held by the user. For example, Jane has the role “Auditor” and that role includes read permission on the “System Audit Log.”
Rule Based Access Control is incorrect because it is a form of MAC. A good example would be a Firewall where rules are defined and apply to anyone connecting through the firewall.
References:
All in One third edition, page 164. Official ISC2 Guide page 187.

Data owners decide who has access to resources based only on the identity of the person accessing the resource.
The following answers are incorrect :
Mandatory Access Control : users and data owners do not have as much freedom to determine who can access files. The operating system makes the final decision and can override the users’ wishes and access decisions are based on security labels.
Sensitive Access Control : There is no such access control in the context of the above question. Role-based Access Control : uses a centrally administered set of controls to determine how subjects and objects interact , also called as non discretionary access control.
In a mandatory access control (MAC) model, users and data owners do not have as much freedom to determine who can access files. The operating system makes the final decision and can override the users’ wishes. This model is much more structured and strict and is based on a security label system. Users are given a security clearance (secret, top secret, confidential, and so on), and data is classified in the same way. The clearance and classification data is stored in the security labels, which are bound to the specific subjects and objects. When the system makes a decision about fulfilling a request to access an object, it is based on the clearance of the subject, the classification of the object, and the security policy of the system. The rules for how subjects access objects are made by the security officer, configured by the administrator, enforced by the operating system, and supported by security technologies Reference : Shon Harris , AIO v3 , Chapter-4 : Access Control , Page : 163-165

There is no such component within kerberos environment. Kerberos uses only symmetric encryption and does not make use of any public key component.
The other answers are incorrect because :
Symmetric key cryptography is a part of Kerberos as the KDC holds all the users’ and services’ secret keys. Authentication service (AS) : KDC (Key Distribution Center) provides an authentication service
Principals : Key Distribution Center provides services to principals , which can be users , applications or network services. References: Shon Harris , AIO v3 , Chapter -4: Access Control , Pages : 152-155.

Access control mechanisms should default to no access to provide the necessary level of security and ensure that no security holes go unnoticed. If access is not explicitly allowed, it should be implicitly denied. Source: HARRIS, Shon, All-In-One CISSP Certification Exam Guide, McGraw-Hill/Osborne, 2002, Chapter 4: Access Control (page 143).

Most of the time users usually choose passwords which can be guessed , hence passwords is the BEST answer out of the choices listed above.
The following answers are incorrect because :
Passphrases is incorrect as it is more secure than a password because it is longer.
One-time passwords is incorrect as the name states , it is good for only once and cannot be reused. Token devices is incorrect as this is also a password generator and is an one time password mechanism.
Reference : Shon Harris AIO v3 , Chapter-4 : Access Control , Page : 139 , 142.

The biometric device that is least user accepted is the retina scan, where a system scans the blood-vessel pattern on the backside of the eyeball. When using this device, an individual has to place their eye up to a device, and may require a puff of air to be blown into the eye. The iris scan only needs for an individual to glance at a camera that could be placed above a door. Source: HARRIS, Shon, All-In-One CISSP Certification Exam Guide, McGraw-Hill/Osborne, 2002, Chapter 4: Access Control (page 131).

From most effective (lowest CER) to least effective (highest CER) are: Iris scan, fingerprint, voice verification, keystroke dynamics. Reference : Shon Harris Aio v3 , Chapter-4 : Access Control , Page : 131 Also see: http://www.sans.org/reading_room/whitepapers/authentication/biometr…