WEBVTT

00:00:01.140 --> 00:00:06.740
Let's take a look at establishing qualitative and quantitative analysis.

00:00:06.770 --> 00:00:10.970
When we think of qualitative assessment or analysis,

00:00:10.970 --> 00:00:13.900
this typically employs a set of methods,

00:00:13.900 --> 00:00:14.460
principles,

00:00:14.460 --> 00:00:20.490
or rules for assessing risk based on non‑numeric categories or levels like low,

00:00:20.490 --> 00:00:21.360
medium, and high.

00:00:21.840 --> 00:00:23.380
Quantitative, on the other hand,

00:00:23.380 --> 00:00:28.690
is going to deal in numerical values typically based off of

00:00:28.690 --> 00:00:31.840
financial lost and financial impacts.

00:00:31.850 --> 00:00:35.620
So qualitative can be described as being subjective,

00:00:35.620 --> 00:00:41.240
whereas quantitative heads toward empirical data use so

00:00:41.240 --> 00:00:44.070
it's more of an objective exercise.

00:00:44.080 --> 00:00:45.970
When we're thinking about qualitative,

00:00:45.970 --> 00:00:50.340
the primary word that you should have in mind is this idea of impact.

00:00:50.350 --> 00:00:52.550
When you think about quantitative,

00:00:52.560 --> 00:00:56.550
the idea should be from the standpoint of impact,

00:00:56.560 --> 00:00:58.360
but based off of financial.

00:00:58.370 --> 00:01:00.370
So qualitative asked the question,

00:01:00.380 --> 00:01:04.700
what would happen to the business function if this was lost?

00:01:04.700 --> 00:01:06.640
Whereas, quantitative asked,

00:01:06.650 --> 00:01:10.900
what would happen on a financial level if this were lost?

00:01:10.910 --> 00:01:15.350
You can normally associate qualitative with categorization and

00:01:15.350 --> 00:01:18.660
quantitative with classification of data.

00:01:18.840 --> 00:01:21.850
When you imagine a qualitative analysis,

00:01:21.860 --> 00:01:23.730
it should be relatively easy,

00:01:23.740 --> 00:01:26.220
even though you are dependent on subject matter experts,

00:01:26.220 --> 00:01:28.900
and when you think about quantitative,

00:01:28.910 --> 00:01:33.190
it is probably going to be quite involved using some

00:01:33.190 --> 00:01:35.550
type of mathematical formulas.

00:01:35.740 --> 00:01:40.450
Here is an example of a qualitative analysis considering a threat.

00:01:40.540 --> 00:01:46.400
If you're familiar with the basic geography of the contiguous

00:01:46.400 --> 00:01:49.220
United States and you think about the West Coast and the

00:01:49.220 --> 00:01:52.250
likelihood of a category five hurricane,

00:01:52.260 --> 00:01:58.060
well at this date at this recording the likelihood is probably low.

00:01:58.140 --> 00:02:03.330
But if I said the West Coast of the United States and a 7.0 earthquake,

00:02:03.340 --> 00:02:06.380
now we're saying, yep, that could be likely,

00:02:06.390 --> 00:02:11.830
and seismologists are predicting that we're overdue for a big one there.

00:02:11.840 --> 00:02:16.510
If the likelihood is high, then consider what the impact would be.

00:02:16.520 --> 00:02:20.700
So qualitative is a really good thing to start with.

00:02:20.710 --> 00:02:22.680
If you end with qualitative,

00:02:22.690 --> 00:02:26.410
it may not end up serving your organization very well because

00:02:26.410 --> 00:02:29.430
you're only dealing with partial impact information,

00:02:29.440 --> 00:02:34.790
operational impact, where now you could move into financial impact information.

00:02:34.800 --> 00:02:38.840
It's also helpful because if I try to focus on all threats,

00:02:38.850 --> 00:02:42.870
then I'm really focusing on nothing, so likelihood becomes a

00:02:42.880 --> 00:02:46.200
cool imperative that we could use for a qualitative to help us

00:02:46.200 --> 00:02:48.660
focus resources in an organization.

00:02:48.840 --> 00:02:54.070
Quantitative analysis says that I am going to consider the financial loss

00:02:54.070 --> 00:02:57.510
that could come from a threat successfully occurring.

00:02:57.520 --> 00:03:02.490
Imagine a scenario where you have a data center that is near a body of water

00:03:02.500 --> 00:03:06.590
and let's say that the water floods its bounds ever so often.

00:03:06.600 --> 00:03:09.940
If we do a quantitative analysis,

00:03:09.950 --> 00:03:15.250
what we are going to experience is probably an actuarial formula

00:03:15.260 --> 00:03:18.110
that helps us to understand the financial impact.

00:03:18.120 --> 00:03:22.150
So we go into that data center and we look at an asset,

00:03:22.160 --> 00:03:25.200
let's say that it is a server rack,

00:03:25.210 --> 00:03:29.650
and maybe it's the cloud service provider that's doing this and

00:03:29.650 --> 00:03:33.530
they know that someone is consuming a private cloud and so that

00:03:33.530 --> 00:03:35.680
this is going to be dedicated services.

00:03:35.680 --> 00:03:39.730
The first thing that they have to do is obtain the asset value.

00:03:39.740 --> 00:03:41.540
So we'll just pick a number,

00:03:41.550 --> 00:03:46.840
and we'll say that the value of this asset is $150,000.

00:03:46.850 --> 00:03:48.180
Now, how is that determined?

00:03:48.180 --> 00:03:51.850
Some people will say is it the cost to replace?

00:03:51.850 --> 00:03:52.890
Is it the opportunity loss?

00:03:52.890 --> 00:03:54.610
Is it the net present value?

00:03:54.620 --> 00:03:56.830
My answer to that is yes.

00:03:56.840 --> 00:04:02.020
It depends on how your organization values resources.

00:04:02.030 --> 00:04:05.100
What is the way in which they look at cash?

00:04:05.110 --> 00:04:11.910
You then would say what is the single loss expectancy of that resource?

00:04:11.920 --> 00:04:13.230
How do I get that?

00:04:13.300 --> 00:04:17.899
I take the asset value that I determined and I multiply

00:04:17.899 --> 00:04:20.589
that by what's called the exposure factor.

00:04:20.600 --> 00:04:24.360
How high am I predicting that that flood comes?

00:04:24.370 --> 00:04:24.710
Well,

00:04:24.710 --> 00:04:28.900
there are actual empirical data that show you how often things

00:04:28.900 --> 00:04:31.890
like flooding should occur in a given area.

00:04:31.900 --> 00:04:35.530
In the United States, there is something known as farmer's almanac.

00:04:35.540 --> 00:04:39.020
When I multiply the asset value by the exposure factor,

00:04:39.050 --> 00:04:41.750
then I get my single loss expectancy.

00:04:42.340 --> 00:04:46.660
The next thing that I have to do is go to work on the annual rate of occurrence.

00:04:47.040 --> 00:04:49.550
How often does this event occur?

00:04:49.830 --> 00:04:53.230
I go back to my empirical data and I see that this

00:04:53.240 --> 00:04:59.950
event happens once every 5 years, 5 into 1 gives me .2.

00:05:00.050 --> 00:05:04.920
Then I'm supposed to take my single loss expectancy results and

00:05:04.920 --> 00:05:09.560
multiply that by my annual rate of occurrence.

00:05:09.570 --> 00:05:10.460
When I do that,

00:05:10.940 --> 00:05:18.560
I then come up with the annual loss expectancy of $3,000 for this 1 situation,

00:05:19.340 --> 00:05:23.020
this 1 event with this 1 resource.

00:05:23.030 --> 00:05:29.520
Think about that $3,000 and what kind of guide can you imagine

00:05:29.520 --> 00:05:32.560
from a business perspective you'd have to follow.

00:05:33.140 --> 00:05:34.270
Keep that in mind.

00:05:34.450 --> 00:05:38.750
We'll get into it in the next clip as we think about this scenario.

00:05:39.640 --> 00:05:43.700
Another type of analysis that an organization could do

00:05:43.700 --> 00:05:46.180
would be a privacy impact assessment.

00:05:46.230 --> 00:05:51.160
This is an analysis of how a personally identifiable information is collected,

00:05:51.160 --> 00:05:53.080
used, shared, and maintained.

00:05:53.090 --> 00:05:58.990
So it demonstrates that program managers and system owners have consciously

00:05:58.990 --> 00:06:03.910
incorporated privacy protections throughout the development lifecycle of

00:06:03.920 --> 00:06:08.260
their system or their program as is related to data.

00:06:08.340 --> 00:06:13.010
They also need to make sure that they communicate in a clear fashion

00:06:13.010 --> 00:06:15.540
with the public about how they handle information,

00:06:15.550 --> 00:06:20.260
including how they would address privacy concerns and safeguard the

00:06:20.270 --> 00:06:23.400
information under use by means of the controller.

00:06:23.480 --> 00:06:24.520
And finally,

00:06:24.530 --> 00:06:29.080
what is the way in which it gets applied when we talk

00:06:29.080 --> 00:06:30.970
about a privacy impact assessment?

00:06:30.980 --> 00:06:35.070
Well it depends on your industry, it could be an access control system,

00:06:35.080 --> 00:06:37.440
it could be a cloud service provider,

00:06:37.450 --> 00:06:42.200
it could be a specific agency that's a government agency,

00:06:42.210 --> 00:06:44.050
a privately held agency.

00:06:44.060 --> 00:06:48.180
Any one of these could actually create a privacy impact assessment.

00:06:48.190 --> 00:06:52.480
You should be thinking about this as a controller or as a

00:06:52.480 --> 00:06:55.160
processor in your environment as well.

00:06:55.340 --> 00:06:58.960
You remember the scenario we use for quantitative analysis?

00:06:59.440 --> 00:07:00.930
Keep that in mind.

00:07:00.940 --> 00:07:03.360
We're going to revisit that in the next clip when we

00:07:03.360 --> 00:07:06.250
talk about risk treatment responses.
