Your question is very useful for an analytic discussion[1], because all electrical measurements with the classical "Bar geometry"[2] are valid only for virtually homogeneous (in directions[3,4], also) bulk materials using the bar (geometry) formed sample(s). Even for the highest symmetry (a cubic symmetry) single crystals, all electrical measurements depend on the (crystal) symmetry of the bar formed sample(s), that should be (always) noted, namely[4]. However, a big (surface) size single crystal graphene (not as the HOPG) sample could be considered as a virtually homogeneous sample having a virtual thickness th, calculated from the out of plane orbitals' distance D (e.g. th ~ 2.D).

1. What do you measure when you measure resistivity? http://aip.scitation.org/doi/abs/10.1063/1.1142958

2. Alternative to "Bar geometry" : Van der Pauw method https://en.wikipedia.org/wiki/Van_der_Pauw_method

3. The phenomenon of anisotropy of the resistance to microcleavage for a carbon steel prestrained in compression http://rd.springer.com/article/10.1007/s11223-007-0071-8?no-access=true

4. Anisotropic Resistance: A Physical Link To Weird Electronic Behavior http://www.science20.com/news_articles/anisotropic_resistance_a_physical_link_to_weird_electronic_behavior-141741

When you measure Hall effect of 2D system the most part of formulas remain in similar form. Biggest difference - you should forget about sample thikness (in fact you can substitute it with "1"), because in truly 2D system there is no motion in this direction. So the Hall constant will be just Rxy/B - i.e. linear slope. If you will carefully (in terms of  units) use this constant you will obtain carrier concentration in m^-2 - which is correct for 2D system. To obtain Hall mobility you should use standart relation but resistivity should be not in m*Ohm, but in Ohm/[] (Ohms per square). Depending on the sample's geometry (Van der Pau or Hall bar mesa) you will have to either solve exponent equation for VdP geometry, or determine geometrical factor for Hall mesa. 

Dear Jan,

For a magnetic field orthogonal to the conduction plane, there is an exact parallelism between planar 3D Hall sensors (e.g. a n-type silicon layer on top of a p-type bulk) and 2D devices. All that you have to do is replacing the product "3D electron density times thickness",    n_3d × t  (unit m^-3 \times m = m^-2) by  the 2D electron density n_s (unit directly in m^-2). This is very generally true because of the isotropy of a second-order tensor (conductivity) in a material with more symmetry than just order 2 symmetry, as stated by Yoannis Samara..

As already stated by LN Oveshnikov:

- the Hall cross sensitivity = 1/(e \ times n_s)  with  unit = m^2/C.

- the 2D resistivity is 1/( e  \ times n_s \ times \ mu)   or simply the "sheet resistance" or "square resistance" of the layer, with unit ohm (called ohm/square in a electrotechnical context). 

- you should trust these 2D physical quantities, and completely forget anything on a "thickness", that is of no relevance here.

Now, when the magnetic field has an in-plane component, all the phenomenology is much more simple in 2D than in 3D. No orbit possible for the electrons out-of the conduction plane, which means a much lower planar Hall effect, i.e. much less sensitivity to the in-plane component of B, see e.g. Morvic and Betko, Sensors and Actuators A 120 (2005) 130–133.

Thank you all for your answers. Much appreciated. This will help me a lot!

The answer of @Vincent Mosser is very useful for discussion. However, an out-of the conduction plane orbits' (OCPO) consideration might be very complex under magnetic field B. We know, already, some chemical[1] influence on OCPO and we might predict that under high B values[2,3], we should expect complex[3,4]  and non linear influences (B-induced strain etc. from OCPO), also.

1. Tunable Band Gap in Hydrogenated Quasi-Free-Standing Graphene http://pubs.acs.org/doi/ipdf/10.1021/nl101066m

2. Two-dimensional monolayer designs for spintronics applications http://onlinelibrary.wiley.com/doi/10.1002/wcms.1259/pdf

3. Magneto-tunable one-dimensional graphene-based photonic crystal  http://aip.scitation.org/doi/pdf/10.1063/1.4870828

4. Band gaps in jagged and straight graphene nanoribbons tunable by an external electric field http://iopscience.iop.org/article/10.1088/0953-8984/27/14/145305/pdf

@ Vincent Mosser I have a question regarding sheet resistance: are we actually measuring sheet resistance in VdP geometry for 2D conductor like we measure resistance in 3D conductor: So did you mean what we measuring in VdP is actually sheet resistivity. sheet resistivity= V/I?