All FST validation studies keep the accuracy of the SFST secret. Till now.
  Home   Claimed Accuracy   Raw Data   Scientific Accuracy   Secret Evidence  

This web site gives the opinions of Dr. Greg Kane. Everything you read here is expressed only as my personal opinion.

What is this data?

ON THIS PAGE, available publicly for the first time, is the data gathered and analyzed in the 1998 San Diego field sobriety test validation study, issued by the NHTSA as: Validation of the Standardized Field Sobriety Test Batter at BACs Below 0.10 Percent. Authors Dr. Jack Stuster and Dr. Marcelline Burns.

THE CLAIM: This data is the basis for the claim: "[O]fficers’ estimates of whether a motorist’s BAC was above or below 0.08 or 0.04 percent were extremely accurate....BAC was above 0.04 percent but lower than 0.08 percent were accurate in 94 percent of the decisions to arrest." [the study's Final Report, page iii]

THE FACTS Show me the graph:
This is a graph of never-before-published data from the NHTSA's 1998 San Diego SFST validation study. Data was gathered by seven experienced alcohol enforcement officers given personal SFST training by Dr. Marcelline Burns, (effectively the inventor of the SFST). Each point on the graph represents one driver in the study: total FST score on the x-axis, Blood Alcohol Concentration on the y-axis.

296 drivers performed at least one of the three components of the SFST.
292 failed – 99%
4 passed – 1% Three of these four drivers were tested by a single officer. The fourth was tested only with OLS.

In other words, seven highly experienced alcohol enforcement officers, personally trained by Dr. Burns, patrolled a major US city for more than five months, stopping and assessing hundreds of motorists with SFSTs. And in all those months, in all those hundreds of tests, only one officer ever completed a single SFST that came back "non-impaired" at the 0.04% BAC level. NHTSA science proves that for six of seven highly experienced DUI patrol officers, every single driver who is able to take the SFST fails the SFST.

At BAC 0.04% six of seven officers failed every driver who could take the test. Their accuracy on innocent drivers was zero percent. Zero percent!

Where did this
data come from?

The NHTSA and SCRI refuse to release this data (or they've lost it, or something).

I got a copy (an Excel file) from researcher Dr. Mike Hlastala, who got it originally from the NHTSA via a FOIA request. Later I got a second copy (an Excel file) from Dr. Jack Stuster, principal author of the study. The two data sets were identical

Show me the graph

Now that you have the raw data, you can do your own statistical analysis of what FST validation studies really show. Here's a taste. Here are graphs with drivers' FST, HGN, WAT, and OLS scores on the x-axis, with the same drivers' BAC on the y-axis.

FST

HGN

WAT

OLS

If you know a person's WAT, OLS and HGN score, do you really know their BAC?

You can download
a copy of this data as
an MS Excel .xls file

Or just copy and paste from this page
The Raw Data set from the 1998 San Diego FST "validation" study
Case Officer Date Time Age Sex HGN OLS  WAT

Officer's
BAC
estimate

Actual
BAC

Disposition
1 4819 24-May 2230 1 1 4 2 6 0.090 0.089 3
2 4236 23-May 2240 1 1 2 2 3 0.070 0.060 1
3 4653 23-May 2244 1 1 6 4 9 0.190 0.230 3
4 3661 24-May 30 1 1 6 4 4 0.140 0.120 3
5 2737 24-May 220 1 1 6 5 3 0.120 0.130 3
6 2737 24-May 120 1 1 2 0 2 0.030 0.020 1
7 4241 24-May 136 1 1 6 7 3 0.110 0.150 3
8 4241 24-May 2217 1 1 6     0.080 0.090 3
9 4241 25-May 136 1 1 6 6 5 0.100 0.110 3
10 3661 25-May 1 1 1 6 3   0.090 0.120 3
11 4486 24-May 2250 1 1 2 1 1 0.010 0.010 1
12 4486 25-May 20 1 1 2 0 2 0.080 0.060 1
13 4236 25-May 2100 1 1 0 1 4 0.120 0.043 1
14 4236 30-May 335 2 1 2 1 1 0.070 0.058 1
15 4653 30-May 255 2 1 2 0 1 0.030 0.040 1
16 2737 29-May 2340 2 2 6 1 5 0.100 0.069 2
17 2737 30-May 335 2 2 6 4 3 0.100 0.140 3
18 4819 30-May 225 1 1 6 6 4 0.120 0.120 3
19 3661 30-May 230 2 1 6 4 3 0.120 0.130 3
20 2737 30-May 250 2 1 2   3 0.030 0.049 1
21 4486 30-May 320 1 1 6 7 5 0.110 0.110 3
22 4486 30-May 225 2 1 2     0.070 0.060 2
23 2737 25-May 155 1 1 6 4 2 0.100 0.120 3
24 4486 25-May 2120 1 1 6     0.300 0.260 3
25 4486 24-May 135 1 1 6 6 4 0.110 0.160 3
26 4486 31-May 235 1 1 6 5 6 0.120 0.120 3
27 3661 30-May 355 2 1 4 2 2 0.050 0.050 2
28 2737 30-May 2158 1 2 6 2 3 0.090 0.080 3
29 4819 1-Jun 120 1 1 4 1 1 0.060 0.050 1
30 4486 31-May 2215 1 2 4 5 3 0.080 0.050 1
31 2737 31-May 2333 1 1 2 3 1 0.030 0.039 1
32 2737 1-Jun 315 1 1 6 4 2 0.090 0.076 1
33 4241 31-May 800 1 1 6 9 1 0.130 0.110 3
34 4236 1-Jun 300 1 1 4 0 0 0.080 0.058 1
35 4236 30-May 2300 1 1 2 0 2 0.040 0.000 1
36 4241 1-Jun 215 1 2 6 5 8 0.130 0.140 3
37 4236 1-Jun 55 2 1 6 1 3 0.100 0.100 3
38 4241 31-May 2330 1 1 6 3 3 0.130 0.140 3
39 3661 2-Jun 30 2 1 6 7 5 0.120 0.100 3
40 4653 1-Jun 130 1 1 6 3 6 0.200 0.240 3
41 4653 6-Jun 10 1 1 6 8 7 0.200 0.220 3
42 4486 6-Jun 2350 1 1 6 5 6 0.150 0.180 3
43 4486 7-Jun 225 1 1 4 8 8 0.110 0.130 3
44 4241 6-Jun 230 1 1 6 5 4 0.120 0.140 3
45 4241 6-Jun 220 1 1 6 3 6 0.140 0.120 3
46 2737 7-Jun 5 1 1 6 3 4 0.120 0.140 3
47 3661 6-Jun 104 1 1 6   3 0.200 0.250 3
48 4241 7-Jun 2220 1 1 6 8 4 0.140 0.120 3
49 4486 8-Jun 100 2 1 6 4 8 0.160 0.150 3
50 4486 8-Jun 5 1 1 4 4 3 0.080 0.080 3
51 3661 31-May 2135 1 2 6 3 6 0.120 0.130 3
52 4236 14-Jun 2150 1 1 4 2 1 0.110 0.050 1
53 2737 13-Jun 340 2 1 2 0 1 0.010 0.032 1
54 2737 13-Jun 303 2 2 4 1 0 0.030 0.046 1
55 4241 13-Jun 242 1 1 6 6 3 0.130 0.100 3
56 4236 13-Jun 220 2 1 6 1 1 0.140 0.160 3
57 3661 6-Jun 2350 1 1 6 4 2 0.160 0.140 3
58 4653 15-Jun 50 1 1 4     0.060 0.050 1
59 3661 8-Jun 2345 1 1 6     0.200 0.220 3
61 4486 14-Jun 35 1 1 6 6 5 0.160 0.320 3
62 4236 14-Jun 50 1 1 6 1 1 0.140 0.150 3
63 4241 15-Jun 320 1 1 6 5 8 0.140 0.160 3
64 4241 15-Jun 135 1 1 6 4 4 0.100 0.130 3
65 4486 23-Jun 145 1 1 6 3 4 0.110 0.110 3
66 4819 23-Jun 115 1 1   1   0.060 0.067 2
67 4819 22-Jun 2228 1 1 2   1 0.040 0.022 1
68 4236 23-Jun 220 1 1 4 6 3 0.140 0.190 3
69 4486 22-Jun 325 1 1 6 5 6 0.230 0.200 3
70 4236 22-Jun 205 1 1 6 4 2 0.100 0.131 3
71 4236 22-Jun 220 1 1 2 3 3 0.100 0.190 3
72 4486 21-Jun 2320 1 1 4 2 2 0.090 0.090 3
73 4241 22-Jun 148 1 1 6 9 3 0.160 0.140 3
74 4236 21-Jun 220 1 1 2 1 1 0.070 0.040 1
75 4236 21-Jun 2100 1 1 2 4 1 0.090 0.048 1
76 4236 22-Jun 145 1 1 4 1 1 0.070 0.050 1
77 4819 22-Jun 222 1 2 6 3   0.130 0.156 3
78 3661 14-Jun 134 1 1 6 4 4 0.140 0.150 3
79 3661 8-Jun 25 1 1 6     0.140 0.120 3
80 3661 13-Jun 2245 1 1 6 3 2 0.180 0.300 3
81 3661 15-Jun 2320 1 1 6 3 4 0.090 0.200 3
82 3661 15-Jun 108 1 1 6 1 7 0.180 0.160 3
83 4236 15-Jun 2220 1 1 6 4 4 0.140 0.200 3
84 4819 20-Jun 2241 1 1 2 5 2 0.000 0.000 3
85 4241 21-Jun 130 1 1 6 4 2 0.110 0.140