Turner-Fairbank Highway Research Center

 
 
 

Highway Safety Information System Guidebook for the Illinois State Data Files

SAS File Formats

Prepared by

Anusha patel Nujjetty
Soumya Sharma
LENDIS Corporation
Turner Fairbank Highway Research Center
Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101-2296

Forrest M. Council
University of North Carolina
Highway Safety Research Center
730 Airport Road
Chapel Hill, NC 27599-3430

Prepared for
Federal Highway Administration
Office of Safety and Traffic Operations Research & Development
U.S. Department of Transportation
Washington, DC 20590

Revised, June 2015

 


 

Introduction

 

7

Details of Major Files

 

9

The Accident Files

 

                                                                                                                9

The Roadlog File   

 

12

Traffic Data

 

17

The Intersectional-location Subfile

 

19

The ÒDeficientÓ Horizontal Curve Files

 

20

Issues Related to Merging Files

 

21

Composite List of Variables

 

25

Accident File

 

 

Accident Subfile

 

33

Vehicle Subfile

 

51

Injured Occupants Subfile

 

71

Roadlog File

 

75

Deficient Curve File

111


Introduction to the Illinois HSIS Guidebook

The Illinois data system includes the following basic files:

¥                Accident data

¥                Roadlog File

¥                Bridge (Structures) File

¥                RR Grade Crossing File                

The first two of these files are currently being captured in the HSIS.  Raw file data are provided to the Highway Safety Research Center where they are retained as backup information.  The documentation (variable listings, definitions, etc.) for these raw files and for the SAS files that are developed from them is available at FHWA offices. The conversion programs developed by HSRC and LENDIS to convert the files into SQL and SAS formats are included in the ProgrammerÕs Guidebook, which is available at the HSIS offices at FHWA.

Beginning in 2004, the HSIS system was converted from SYBASE relational database to ORACLE relational database for internal use. This ORACLE database stores the data received from Illinois and other States, and the data files for a given State are linked and manipulated using SAS code.  However, this conversion from the original SAS-based system to the ORACLE relational system is somewhat transparent to the end-user of the data since the output files produced by ORACLE for modeling and analysis will be SAS formatted.  As in the past, we have continued to produce SAS format libraries for each of the variables in each of the files.  Because it is envisioned that the majority of analyses will utilize these SAS files and formats, this Guidebook will concern these SAS files - their formats, completeness, and quality. However, researchers requesting data from HSIS can request the output in various formats such as SAS, Microsoft Excel¨ and Access¨, dBase, ASCII, etc.

It is noted at this point that the "raw files" received from Illinois are less "raw" than in other States since the data do not come directly from police accident report forms or from basic roadway inventory files.  Instead, Illinois has developed their own safety information system, which includes a number of data edits and quality checks.  The "raw" files that we are receiving for HSIS are those files that have already undergone this series of checks and edits.

As noted above, the SAS accident data is in three separate subfiles, the first containing the basic accident information on a case-by-case basis, and then separate files containing information on vehicles and occupants in each accident case.  In the 1985-91 files, the Occupant subfile only includes data on injured occupants. From 1992 onwards, the occupant subfile contains both injured and uninjured occupants.  The vehicle and injured-occupant data can be linked to the basic accident data for specific cases using the accident report number and vehicle number.  The accident subfile can be linked to the Roadlog file using three common variables -- county, route, and milepost.

Unlike an Accident File record that is referenced to a point on the roadway, each record on the Roadlog File contains information on a homogeneous section of roadway (i.e., a stretch of road which is consistent in terms of certain characteristics), with each new section being defined by a new beginning reference point.  Each record on this Roadlog File contains current characteristics of the road system including surface type and width, shoulder and median information, lane information, etc.

Due to established priorities of effort, HSRC staff did not work with the Bridge File or the Railroad Grade-Crossing File.  As is the case with most States, the Bridge File data contains information on bridge structures across the State.  The data is considered quite accurate since it is based on the Federal bridge inventory.  In a similar fashion, the Railroad Grade Crossing File is a file containing information on all grade-crossings in the State, and is prepared and maintained according to FHWA requirements.

Details of the three Accident subfiles and the Roadlog file are presented in the following section.

DETAILS OF MAJOR FILES -                                                                                                                                                                  

The Accident Files

All police departments collect the accident data statewide on a standard form.  Prior to 1992, crashes were reported if they involved personal injury or total property damage of $250 or more.  The property damage threshold increased to $500 on January 1, 1992. Later, threshold was increased to $1500 on January 1, 2009 if all the drives involved in a crash have insurance. If not, the threshold remains $500. The form is sent to the Division of Traffic Safety where coders/editors first match motorist reports with police reports, then look up a location code from a series of maps and "locate" each crash, and then code the basic accident data into the files. (The location code also triggers a computer table "look-up" which allows Illinois to attach a variable related to the Federal functional class to the accident data.  In other States, such a variable would only be found on the roadway inventory file since the investigating officer would not be able to distinguish the various system types.)

The Illinois accident report form changed significantly in 1993.  This new "self-coding" form resulted in new variables, changes in codes for existing variables, and changes in the manner in which the data is processed.  As expected, the data for the "transition year" (i.e., 1993) are perhaps not as accurate as the data for other years.  However, Illinois staff has conducted extensive data edits and recodes to correct as many problems as possible.  In most cases, the data received by HSIS were recoded to "old formats" by Illinois for the 1993 file. However, additional codes were added at times. In 1994, new codes will be used by Illinois.  To continue to allow use of multiple years of data, HSIS staff has modified the new Illinois codes such that they can be integrated with the old HSIS codes. These changes are noted in the SAS format sections that follow.

Approximately 300,000 accidents per year are reported in Illinois.  The HSIS data set contains the subset of 1985-2010 accidents that occurred on the State-inventoried system.  Almost all of these accidents can be linked with the Roadlog file.  This data set includes between 105,000 and 144, 512 accidents per year over the twenty two-year range and between 205,000 and 275, 643 vehicles records per year for the same range. There are between 27,000 and 52,000 occupant records per year during 1985 to1991, and between 74,000 and   100,129 records per year during the 1992-2010. This increase is due to inclusion of the uninjured occupants from 1992 onwards as discussed earlier.  In general, the number of linkable accidents and related vehicles has decreased over time.

For consistency with other HSIS states, pedestrian/equestrian/pedalcycle accidents (which are a subset of the Illinois raw accident file) have been merged into the Vehicle subfile.  A "vehicle" on the accident form which is actually a pedestrian, equestrian, or pedalcycle can be identified by VEHTYPE=98 in the variable related to "Type of Vehicle." The remaining 100,000 to 200,000 "unlinkable" accidents each year are predominantly those occurring on city streets or minor township roads.  These roadways are inventoried to a lesser extent than are the higher order roadways, but this lower level inventory has not been obtained for HSIS use since the accident coders do not attach a specific location reference code to the accident reports for such locations.

In addition, because no inventory data exist for "toll roads" or ÒrampsÓ in the Roadlog File, these accidents (which were included in the HSIS accident files for 1985-87) cannot be linked with roadway data for analysis purposes.  Since the majority of the HSIS analyses involve linkage with roadway data, a decision was made to delete these toll road accidents from the HSIS files beginning with 1988.  In addition, a second decision was made to delete all ÒrampÓ accidents from the files beginning with the 1993 files.  This results in an approximate 10 percent decrease in the total number of accidents (and vehicles and occupants) in the files between the 1992 data and the 1993 data.

As with other States, not all accidents on the file are reported by police.  Approximately 4.8 percent of the records on the original 1985-87 files were driver reports which have no corresponding police report, and thus the data from the driver report is punched directly into the system.  An additional 11.2 percent of the early files are "desk reports" filed by the State Police, city police or county sheriffs.  These desk reports are not the result of a direct investigation by an officer on the scene, but instead are based on information provided to an officer at a police headquarters.  Thus, both the desk reports and the driver reports would be expected to be somewhat less accurate than the "normal" reports filed by investigating officers on the scene.  Driver reports would be assumed to be even less accurate due to expected "self-reporting" biases in variables related to fault, driver condition, safety belt usage, or other variables related to vehicular laws.  Examination of the data did indeed indicate far more "unknown" and "not stated" codes in these driver reports.  Due to this possible bias, the large size of the sample even without these reports, and the subsequent elimination of these reports from all HSIS analyses conducted during the 1988-89 period, the driver and desk reports have been eliminated from 1988 and later Accident Files.  It is strongly recommended that the remaining 1985-87 driver and desk reports be eliminated from all analyses.

Approximately 70 percent of the accidents on the linkable file are property damage only and 0.5 percent is fatal accidents. (Single-variable tables presented in a later section of this documentation provide further details of these breakdowns.)  Approximately 82 percent of the accidents are multi-vehicle in nature, while the remaining 18 percent are single-vehicle accidents of one type or another. When reporting agency was examined, it appears that the Illinois files are slightly more biased to urban/suburban crashes than are some of the other HSIS states.

An assessment of the completeness and accuracy of the data is based on conversations with Illinois staff, use of the data in prior HSIS analysis efforts, and a series of single-variable tabulations run each year for key analysis variables.  These quality-control runs allows the HSIS staff to examine both the percent uncoded for each variable and changes across time in the individual codes within each key variable.  These yearly runs are performed on all the variables.

These runs have consistently indicated that almost all of the variables in the three subfiles have very few uncoded cases and very few error codes.  The only significant differences noted across the coded categories appear to be the result of the new report form in 1993.  Where high numbers of uncoded cases or inconsistencies in codes are found, a "NOTE" has been included under the pertinent variable in the later SAS format sections.

In addition to the quality-control checks noted above, in order to further check the accuracy of some of the accident variables, a series of comparisons were made of variables which should have been somewhat similar on the accident subfiles according to their definitions (both within the same subfile and across subfiles).  These single-variable tabulations were also compared to HSIS staff knowledge of the North Carolina accident files to see if major (unexplained) discrepancies existed. In general, the comparisons indicated the files were quite accurate.  For example, the rural and urban breakdown in the "Federal Functional Class" variable matched quite well with the rural and urban breakdown in the "Traffic way Class" variable.  The counts under the variable related to "Total Injuries" matched quite well with the accident severity variable, and counts from the collision type variable matched well with what would be expected from the "Number of Vehicles" variable.  Again, there are a few cases in which variables are either less than totally consistent with other variables or have changed across time (e.g., head-on and sideswipe codes under "Type of Collision" appear to indicate opposing vehicles directions immediately prior to impact rather than their initial direction.) Again, where found, a "NOTE" has been included in the SAS formats.

However, in general, analysis of the Illinois accident, vehicle, and occupant subfiles indicated that these files are quite accurate in almost all cases, are quite internally consistent, and, with few exceptions, have very few "not stated" cases.  The only major inconsistencies are due to changes in codes in the 1993 data and the fact that uninjured occupants are included in the Occupant Subfile for 1992 and 1993. For more details of the changes in codes from 1993 onwards, please refer to the format files.

[SS2] The Roadlog File

As noted above, the Roadlog file contains current characteristics of the road system.  It contains information on homogeneous sections of highway for approximately 17,247.27miles of roadway, with approximately 2,100 miles of Interstate, 10,100 miles of other primary roadway, and 5,000 miles of secondary, county and township roads.  Currently, there are twenty one Roadlog files in the HSIS system -- 1987-2010.  Beginning in 1997, IL DOT started sending us new roadway files based on a new referencing system. When the mileages by different roadway types were checked against the 1994 data it was found that there was some increase in the mileage for various categories for 1997, but the mileage remained consistent between 1997 and 2010 data. The IL DOT indicated that in these new files there are some coinciding routes that were causing an increase in the total mileage due to double counting. However, the crashes are always posted to the higher order routes if two routes coincide. Hence it was decided to eliminate the lower order coinciding routes using CON_CURR variable. This elimination of roadway segments made the total mileage comparable to previous years. These new roadlog files contain additional data on intersections, horizontal and vertical curves that is different from the previous year files. More detailed explanation of these new variables is given below. Table 1 below provides a categorization of all two-way paved mileage in the 2010 data.

It should be noted that the definition of "homogeneous section" is based on fifteen different variables -- number of lanes, access control, one-way/two-way operation, shoulder width, shoulder type, median width, median type, parking lane width, ADT, speed limit, vertical curve length, horizontal curve radius, and three variables related to political boundaries and city size. Thus, anytime one of these variables changes, a new section is started.  However, it is also noted that a new section is not begun with changes in other variables -- pavement width and percent trucks, for example. The inventory file is updated yearly based on inventories performed by district field personnel.  When a new section of roadway is planned, a "blank" record is put into the file indicating that a roadway section will be added, but no data is entered at this point.  Once the roadway is completed and opens for traffic, it is inventoried by the field personnel.  The characteristics of the completed road are compared to those on the plans, and a listing of the true characteristics is sent back to headquarters to be entered in the inventory file.

Table 1 HSIS roadway mileage by roadway category (2010 data).

 

Roadway Category

Mileage

Urban freeways

981.61

Urban freeways < 4 Lanes

15.65

Urban multilane divided non-freeways

1,342.92

Urban multilane undivided non-freeways

899.66

Urban 2 lane highways

2,309.36

Rural freeways

1,382.31

Rural freeways < 4 lanes

35.82

Rural multilane divided non-freeways

325.16

Rural multilane undivided non-freeways

39.25

Rural 2 lane highways

9668.33

Other

247.2

Total

17,247.27

 

A similar updating procedure is carried out when improvements (such as 3R improvements) or modifications of the roadway are carried out.  When a large-scale modification such as a lane widening or curve flattening is planned for a section of road, district engineers make note of the fact that the roadway inventory will have to be changed as soon as the road is completed.  When the modification is completed, the field personnel inventory the new roadway; send the revised data to headquarters, and the data is entered into the inventory file. The file is then completely updated at the end of each year.  (It is noted that the updated file is usually available around April 1 of the following year.)

The file contains approximately 80,000 to 180,364 records during the period 1987-2010, each representing a section of "homogeneous" roadway where characteristics remain constant.  Thus, the average section length is about 0.15 mile.   Approximately 60 % of the mileage is in non- municipal (rural) areas, and 70 % are two-lane roadways.

Note that prior to 1997, horizontal curvature information was captured on the roadlog file rather than on a separate "curvature" file.  (As will be described below, a separate curvature file exist for 1997 and later years.) In those earlier years, curve section beginning and end points are defined somewhat differently from what might be expected.  More specifically, when curvature changes within a given section (say from a tangent to some degree of curve), the section is ended (i.e., a new section is begun) at the Point of Intersection of the tangents (PI) of the horizontal curve, not at the beginning of the curve.  The "old" section (which actually contains one-half the curve) contains no information on curvature variables.  The new roadlog section, which begins at the PI, contains all of the descriptive information related to the curve such as deflection angle, direction, radius and length. In some sense, if one wished to be able to define exactly where the curve existed on a particular piece of highway, one would have to go to the reference point which exists at the mid-point of the curve and then "back-track" for half the length of the curve back into the preceding section.

Fortunately, this does not raise great problems if one is simply trying to link accidents with their respective curvature data since the DOT coders who "locate" the accidents always code all accidents on a given curve to the reference point for the PI. Thus, accidents can be linked to their respective curve information.  The only exception that might cause some problems would be if an intersection fell within the curve itself.  In this case, accidents, which the coders judge to be related to the intersection, would be coded to the intersection rather than to the PI of the curve. In summary, for years prior to 1997 information on horizontal curves is available in the roadlog file. From 1997 onwards, this information is available in a separate curvature file.

However, of significant importance from an analysis perspective is the fact that the only curves which have complete data on the file are those "potentially substandard" curves which have radii equal to or less than 2291.83 feet (approximately a 2.5 degree curve).  Detailed data are not collected on flatter curves, although partial data (such as deflection angle or direction) may appear on the file if the district engineer wishes to have a "curve marker" there.  It is also the case that information on curves is only collected in rural areas and is not collected on Interstate roadways where all curves are assumed to meet federal standards.  However, there is some information on curves on Interstates, when a district engineer feels that the curve is sub-standard. However, this is neither a complete census nor a random sample and care should be exercised while using this data. Thus, it would appear that the most consistent curvature data would be on those curves where a radius is greater than 2291 feet, and the analyst must be aware that the range of curves available for analysis does not include the flatter (less than 2.5 degree) curves.

For all years, information on vertical grade and vertical curvature is also included on the Roadlog File.  However for years prior to 1997 no beginning and ending milepost information is available. Hence linking crashes to vertical curves for these years is not possible. Also, similar to the horizontal curve data, preliminary analyses with the vertical grade and curve variables and subsequent conversations with Illinois staff indicated that these vertical variables are only coded in those cases in which the engineering staff considers the existing vertical curve to be "substandard" in nature.  Thus, the vertical grade data is neither a full sample nor a random sample of the roadway, and care must be taken in using it in analyses.

As with the accident files, the assessment of accuracy and completeness of the Roadlog File data is based on input from Illinois staff, prior HSIS analyses with the data, and a series of single- variable tabulations for all variables in the file which are run and reviewed each time a new annual file is entered into the HSIS. In addition, traffic variables concerning the AADT and percent trucks, and alignment variables concerning horizontal curve deflection angle, radius, and direction and vertical curve length and entering/leaving grade are included.

The quality control runs consistently indicate that most of the variables contained very few uncoded or obviously miscoded values -- only one to three percent in most cases.  There are some cases where a "zero" value may well be both a valid value and may indicate uncoded data (such as with surface condition rating), but even here the percent of sections coded as zero is less than one percent of the file.  Where higher percentages of uncoded data or error codes have been found, a "NOTE" is included in the SAS format.

Further assessment of the accuracy of the data included comparison of the limited number of similar variables in the file.  These comparisons, subsequent HSIS analyses with the files, and the update procedures described above lead to the conclusion that the roadway inventory data is quite accurate.

Two new variables, RODWYCLS and MVMT, have been created by HSIS staff in the roadway segment file of each of the HSIS states. The RODWYCLS (Roadway Class) variable is based on the combination of rural/urban, access control, number of lanes and median type variables. This variable classifies each roadway segment into one of ten roadway types described in the later ÒFormatÓ section. This variable is also included as an accident-file variable by matching each crash to its corresponding roadway segment. The MVMT variable (Million Vehicle Miles of Travel) is calculated for each segment in the roadway file by multiplying the segment length, AADT and 365 days in a year, and dividing by one million. Both these variables were created in response to inquire from data users, whose most frequent questions have concerned either crash frequencies or rates (per MVMT) for one or more of these roadway classes.

HSIS staff has stopped labeling shoulder type and shoulder width variables as left and right shoulder type and shoulder width respectively from 1997 onwards.  Instead, they are labeled as inside and outside shoulder types and shoulder widths. Further details along with the new variable names are given in the FORMAT section of this guidebook.  This change in variable labeling is based on the definition provided in the IRIS manual if IL DOT.

Traffic Data

As indicated earlier the Roadlog File contains information on AADT, percent trucks for 1990 and earlier, and commercial vehicle AADT for 1991 and later files.  These data are developed in Illinois' traffic volume counting program, and are based on a combination of permanent counters that count traffic 24-hours each day for 365 days each year and a series of short-term "coverage" counts conducted each year.  Before 2001, Illinois had 49 automatic traffic recorders (ATRs), of which 21 were capable of collecting counts by vehicle class in accordance with FHWA's Scheme F.  The ATR locations on the five different classes of roadway included seven on rural Interstate roadway, six locations on urban Interstate locations, 12 locations on other rural roadways, 19 locations on other urban routes, and five locations on ÒrecreationalÓ routes. From 2001, Illinois has 85 ATRs, of which 36 are capable of collecting counts by vehicle class in accordance with FHWA`s Scheme F. The ATR locations on the five different classes of roadway include 14 locations on rural interstate, 5 locations on urban interstate, 14 locations on other rural roadways and 52 locations on other urban routes.

 

In addition to the ATR data, short term traffic counts on Interstate and primary highway systems are done on a 2 year cycle. Before 2000, during even-numbered years, portable counter devices were deployed in combination with pre-established in pavement loop detectors.  Typically, the counter devices were deployed during one week of the year at any given site.  Short counts (e.g., 24- or 48 hour counts) were collected on Monday through Thursday only. It should be noted that a sample of Interstate sections was counted one week out of every four months.  During odd-numbered years, the Illinois DOT conducted a comprehensive interchange ramp counting program on State Highways.  These ramp counts were used to supplement ADT data for sections where the State did not have monitors (i.e., counter devices).  In total, it was estimated that approximately 96 percent of the primary system is covered during each two-year cycle. From 2000, Illinois started using magnetic lane counters for 24-hour counts during every odd-numbered year. The magnetic lane counters provide volume as well as vehicle classification by length in three categories (passenger vehicles, single-units, multi-units).  Using the magnetic lane counters allow IDOT to collect mainline data for all State primary routes, except the interstates in Chicago area.  The Chicago area interstates are counted with loops by the Expressway Surveillance program and the Illinois Tollway Authority.  Collecting the mainline data eliminates the need for ramp counting.  Short counts (e.g., 24- or 48 hour counts) are collected on Monday through Thursday only. In total, an estimated 99 percent of the primary system is covered during each two-year cycle.

 

For other non-primary roads (i.e., the "off" marked route system), Illinois collects 24-hour coverage counts in approximately 20 percent of the counties once every five years. However, the Northeast Counties are done every four years.  With the exception of Cook County, which is also on a four-year cycle, urban areas within counties are counted on a five-year statewide cycle.

Before 2001, additional vehicle classification counts are conducted on HPMS sections. These were made at 300 locations over a three-year cycle (i.e., approximately 100 each year) to form a representative distribution for the State.  The 300 HPMS sections use for classification have been eliminated from 2001 onwards.  Instead of using HPMS sample locations, the use of the magnetic lane counts provides vehicle classification by length for the complete State Primary system every two years.  Rather than sampling, system-wide truck VMT based on actual counts is used.

Finally, the districts often have need for additional traffic data.  Consequently, when requested, the State collects 12 hour turning movement counts at intersection and other "special" traffic data to satisfy these needs.

To convert the short-term coverage counts to AADT, Illinois applies adjustments for seasonal differences in the daily traffic.  For seasonal corrections, each coverage count location is assigned to one of the five categories of roadway where permanent counters are located as defined above.  The seasonal factors are based on averages from all ATRs in that group.

When a road section is not counted during a given year, growth factors are developed and applied to the most recent prior year's count. Average growth factors are created each year for each functional class of roadway using ATR data and data from adjusted short counts for the current year. The growth factor applied to a particular uncounted section is based on its functional class.  For sections where no prior AADT exist, AADT/mile averages by functional class are developed and then used in order to "fill in" the AADT's.

Finally, it should be noted that the percentages of truck-related "Heavy Commercial VolumesÓ included "two-axle trucks with six or more tires plus multi-axle vehicles", prior to 2001. Thus, while pick-ups and vans were excluded, this combination did include single trucks, tractor- semi combinations and buses.  Thus, it couldnÕt be considered a count of just the multiple unit (tractor-trailer) trucks that were found on the roadway system.

After 2001, IllinoisÕ implementation of the vehicle length for truck classification instead of axle classification was approved by FHWA and confirms the new Traffic Monitoring Guide. Illinois worked closely with FHWA during the time of research and testing on the different equipment that was looked at to solve Illinois needs.  While the permanent ATR locations still collect the Scheme F categories, Illinois only uses the three categories (PV, SU, MU) for their publications and in-house use.

The Intersection-Location Subfile

(Note that in 1997, Illinois stopped providing the intersection data as described in this section. For 1997 and later years, each roadway segment contains three sets of variables to indicate location, type and orientation of intersections. Further details are provided below each of the variables in the FORMAT section of this guidebook for roadlog file.)

The Illinois base roadway inventory file contains two types of records.  First, homogeneous sections are defined by unique beginning and ending mileposts, with the two milepoints never being equal.  Second, an intersection-location record is included for each intersection, which falls in a given homogeneous section.  Thus, if three intersections fall within a given section, three records with mileposts between the original beginning and ending mileposts will be provided.  These intersection- location records are defined by having the beginning and ending mileposts equal to each other.

In building the HSIS files, we have separated all intersection records into a separate file, with one such file corresponding to each yearly Roadlog File.  Each record on this file contains the milepost (reference point) and all variables appearing on the Roadlog section file, and the data included refer to the section of roadway downstream from the intersection.  Unfortunately, this file does not contain desirable intersection-related variables such as type of intersection (e.g., T or cross), entering traffic volumes, or traffic control type (although this latter variable may be available after 1989). However, this file can be used in two ways.  First, if one were to desire to do an analysis of "clean" sections without intersection-related accidents being part of the analysis, the Intersection- Location subfile could be merged with the appropriate sections in the Roadlog file, and all accidents, which fall at the intersection milepoints, could be deleted.  (Illinois codes all intersection-related accidents to the specific milepost of the intersection, even if they occur upstream or downstream.) Second, as noted above, if one wished to analyze variables which do not cause breaks in sections (such as surface width or percent trucks), then it would be possible to divide the existing sections into shorter sections, with a break occurring at each intersection.  These shorter sections could then be merged back into longer sections with breaks only on variables of interest.  Some programming has been developed to do such merging.

The ÒDeficientÓ Horizontal Curve Files

As indicated above, curvature information on Òdeficient curvesÓ is captured on the Roadlog File prior to 1997.  From 1997 onwards, HSIS Staff has extracted new curvature variables from the Roadlog File and created a horizontal curvature file. As in earlier years, these files continue to contain information on ÒdeficientÓ horizontal curves.  IL defines a curve to be ÒdeficientÓ if its radius is less than 2291.83 feet (>= 2.5 degrees). Detailed data are not collected on flatter curves, although partial data (such as deflection angle or direction) may appear on the file if the district engineer wishes to have a "curve marker" there.  It is also the case that information on curves is only collected in rural areas and is not collected on Interstate roadways where all curves are assumed to meet federal standards. However, the engineering staff has recorded some curves on the interstate system, which are captured in the data. As stated above, these curves on interstate are neither a random sample nor a complete census. The majority of reported curves occur on U. S. and State maintained routes.  An annual file is available for years 1997 to 2010. On the average it contains 3,400 curves covering approximately 450 miles.  Based on the values of radius of curve, degree of curve has been calculated and added to the file.  Thus, this file includes information on the degree of curve, the direction of curve, length of curve and radius of curve.

Issues Related to Merging Files

As noted above, the Illinois accident data are subdivided into three subfiles -- accident, vehicle and occupant.  These subfiles can be linked together using the "case number" variable (i.e., CASENO), which is present in each of the three files.

This variable includes the accident year.  When linking the Occupant subfile, the additional linking variable "vehicle number" (i.e., VEHNO) must match so that the occupants are associated with the vehicle in which they were traveling.  To link the Vehicle subfile with the Accident alone, first sort both subfiles by case number.  To link the Occupant file with the other two subfiles, first sort both the Vehicle subfile and Occupant subfile by case number and vehicle number. Next sort the Accident subfile by case number.  Alternatively, the separate subfiles can linked by specifying an SQL JOIN operation with the constraining condition that case number and vehicle number from each table are equal.  SQL processing does not require the data to be presorted and the output will not be in any particular sort order unless ORDER BY is specified.

The Accident subfile can then be linked with the Roadlog File using information related to county, route prefix/type (e.g., I, US, MN, etc.), route number and milepost on the route.  The actual linkage variables on the Accident file, which are used in the merging operation, are CNTYRTE and MILEPOST.  The linkage variables on the Roadlog File are BEGMP, ENDMP and CNTY_RTE. The CNTYRTE and CNTY_RTE variables are computed linkage keys, which are created by combining variables, related to county (COUNTY), route prefix (RTE_PREF), and route number (RTE_NBR).

To prepare the Accident subfile for linking with the Roadlog File using a SAS data step process, the analyst must sort both the Accident and the Roadway File into location order by CNTYRTE and MILEPOST on the Accident file and by CNTY_RTE and BEGMP on the Roadlog File.  For the alternative SQL join, the analyst must specify an exact match on CNTYRTE and CNTY_RTE from the Accident and Roadlog files, and a range match where MILEPOST occurs between BEGMP and ENDMP.  (Programs to accomplish this merging are documented in the HSIS Programmer's Guidebook, available at FHWA.) Finally, where appropriate and possible, a format, which defines categories within a given variable, has been developed for HSIS SAS variables.  These categories are shown in the pages below.  These formats have been saved in a format library, which can be provided to the user.  As a naming convention, the "format name" is the same as the variable name; with the only exception being for certain character variables (in contrast with numeric variables).  More specifically, a SAS format name has to be preceded by a "$" if the variable is character in nature.  There is an 8-character length limit on both variable name and format name. In cases where the variable name is already eight characters in length, the addition of the preceding "$" would make the format name one character too long. In these cases, the format name is the same as the variable name except the final character of the variable name is dropped.

 


SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

AADT

ANNUAL ADT

Roadlog

NUM

79

AADT_YR

YEAR OF ADT

Roadlog

CHA(4)

79

ACC_DATE

DATE OF ACCIDENT

Accident

NUM

36

ACCESS

ACCESS CONTROL

Roadlog

NUM

79

ACCTYPE_POST_93

TYPE OF COLLISION

Accident

NUM

35

ACCTYPE_PRE_93

TYPE OF COLLISION

Accident

NUM

35

ACCYR

ACCIDENT YEAR

Accident

NUM

36

ACTION

ARREST

Vehicle

CHA(7)

53

ADMINHWY

ADMINISTRATIVE HIGHWAY SYSTEM

Roadlog

CHA(1)

79

AGE

OCCUPANT AGE

Occupant

NUM

72

AGENCY

INVESTIGATING AGENCY

Accident

NUM

36

AIRBAG

AIRBAG DRIVER

Vehicle

CHA(1)

53

ALIGN_CODE

ALIGNMENT

Accident

NUM

37

APPR_NBR

APPURTENANCE NUMBER

Roadlog

NUM

79

AT_FAULT

AT FAULT

Vehicle

CHA(2)

53

AVAI_ROW

AVAILABLE RIGHT OF WAY

Roadlog

CHA(1)

80

BADGE

BADGE CODE

Accident

NUM

37

BEAT_CDE

BEAT CODE

Accident

NUM

37

BEGMP

BEGIN MILEPOST

Roadlog

NUM

80

BEGMP

BEGINNING MILEPOST

Deficient

CURVE NUM

112

BUILD_BY

BUILT BY

Roadlog

CHA(1)

80

CASENO

ACCIDENT CASE NUMBER

Accident

CHA(11)

37

CASENO

ACCIDENT CASE NUMBER

Occupants

CHA(11)

72

CASENO 

ACCIDENT CASE NUMBER 

Vehicle

CHA(11)

53

CAUSE1

CONTRIB FACTOR 1

Accident

CHA(2)

37

CAUSE2

CONTRIB FACTOR 2

Accident

CHA(2)

37

CITY

CITY OR TOWNSHIP

Accident

NUM

38

CITY_TWNSHIP_FLG

CITY/TOWNSHIP FLAG

Accident

CHA(1)

38

CLS_TFWY

CLASS OF TRAFFICWAY

Accident

NUM

38

CNTY_RTE

COUNTY ROUTE NUMBER

Deficient Curve

CHA(7)

112

CNTY_RTE

COUNTY ROUTE NUMBER

Roadlog

CHA(7)

80

CNTYRTE

COMPUTED LINKAGE KEY

Accident

CHA(7)

39


SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

COL_TYPE

COLLISION TYPE

Vehicle

CHA(2)

53

COMM_VEH

COMMERCIAL VEHICLE

Vehicle

CHA(1)

53

COMM_VOL

COMMERCIAL VOLUME

Roadlog

NUM

80

COMMDATE

DATE

Roadlog

CHA(4)

81

COUNTY

COUNTY

Accident

NUM

39

COUNTY

COUNTY

Roadlog

NUM

81

CRSH_LAT

CRASH LATITUDE

Accident

CHA(3)

39

CRSH_LONG

CRASH LONGITUDE

Accident

CHA(3)

39

CRSH_X_CORD

CRASH X COORDINATE

Accident

CHA(8)

39

CRSH_Y_CORD

CRASH Y COORDINATE

Accident

CHA(8)

39

CTY_CLS

CITY CLASS CODE

Accident

NUM

39

CURB1

CURB TYPE

Roadlog

NUM

83

CURV_CUT

CURVE CUT

Roadlog

CHA(1)

83

CURV_LGT

CURVE LENGTH

Deficient Curve

 NUM

112

CURV_RAD

CURVE RADIUS

Deficient Curve

 NUM

112

CURV_RAD

CURVE RADIUS

Roadlog

NUM

83

DAM_OTHR

PROPERTY DAMAGE OTHER THAN VEH

Accident

NUM

40

DEF_ANGL

DEFLECTION ANGLE

Roadlog

CHA(7)

83

DEG_CURV

DEGREE OF CURVATURE

Deficient Curve

 NUM

112

DIR_CURV

DIRECTION OF CURVE

Deficient Curve

CHA(1)

112

DIR_CURV

HORIZONTAL CURVE DIRECTION

Roadlog

CHA(1)

83

DIR_TRVL

DIRECTION OF TRAVEL

Vehicle

NUM

53

DIST

DISTRICT

Accident

NUM

40

DISTRICT

ILL DISTRICT

Roadlog

NUM

84

DIVIDED

TRAFFICWAY DESCRIPTION

Accident

NUM

40

DRV_ACTN

DRIVER ACTION

Vehicle

CHA(2)

54

DRV_AGE

DRIVER AGE

Vehicle

NUM

54

DRV_BAC

DRIVER ALCOHOL PERCENT

Vehicle

NUM

54

DRV_BAC2

2ND SOBRIETY TEST RESULTS

Vehicle

CHA(2)

55

DRV_CLAS

DRIVER CLASS

Vehicle

CHA(4)

55

DRV_COND

DRIVER CONDITION NEW

Vehicle

CHA(1)

55

 

SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

DRV_DOB

DRIVER BIRTH DATE

Vehicle

CHA(8)

55

DRV_EJCT

DRIVER EJECTION

Vehicle

CHA(1)

56

DRV_IMAG

IMAGE NUMBER

Vehicle

NUM

56

DRV_INJ

DRIVER EXTENT OF INJURY

Vehicle

NUM

56

DRV_LST

DRIVER LICENSE STATE

Vehicle

CHA(2)

56

DRV_REEL

REEL NUMBER

Vehicle

CHA(4)

56

DRV_REST

DRIVER RESTRAINT USAGE

Vehicle

NUM

56

DRV_RPT

DRIVER REPORT

Vehicle

NUM

57

DRV_SEX

DRIVER SEX

Vehicle

NUM

57

EJCT

OCCUPANT EJECTION

Occupant

CHA(1)

72

END_RTE

END OF ROUTE

Roadlog

CHA(1)

84

ENDMP

END MILEPOST

Deficient Curve

NUM

113

ENDMP

END MILEPOST

Roadlog

NUM

84

EXST_ROW

EXISTING RIGHT OF WAY

Roadlog

NUM

84

F_INVLOC

FIRST INVOLVEMENT LOCATION

Vehicle

NUM

57

FAUL_HGHT

FAULT HEIGHT

Roadlog

NUM

84

FED_AID

FEDERAL AID (IN LIEU)

Roadlog

NUM

84

FED_CLAS

FEDERAL CLASSIFICATION

Accident

NUM

40

FIRE

VEH FUEL LEAKS AND FIRE

Vehicle

NUM

58

FLD_NAM1

FIELD REF NAME 1

Accident

CHA(3)

41

FLD_NAM2

FIELD REF NAME 2

Accident

CHA(3)

41

FLD_NBR1

FIELD REF NBR 1

Accident

NUM

41

FLD_NBR2

FIELD REF NBR 2

Accident

NUM

41

FLD_TYPE

FIELD REF TYPE

Accident

NUM

41

FRST_INV

FIRST INVOLVEMENT

Vehicle

NUM

58

FUNC_CLS

FUNCTIONAL CLASS

Accident

CHA(1)

41

FUNC_CLS

FUNCTIONAL CLASS

Roadlog

NUM

85

HIT_RUN

HIT AND RUN

Accident

CHA(1)

42

HOR_BEG

HORIZONTAL CURVE BEGINNING MILEPOST

Roadlog

NUM

85

HOR_BEGMP

HORIZONTAL CURVE BEGIN MILEPOST

Roadlog

CHA(5)

85

HOR_END

HORIZONTAL CURVE ENDING MILEPOST

Roadlog

NUM

85

HOR_ENDMP

HORIZONTAL CURVE END MILEPOST

Roadlog

CHA(5)

86

HOUR

TIME OF ACCIDENT

Accident

NUM

42

HPMS_IND

HPMS INDICATOR

Roadlog

CHA(1)

86


SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

HPMS_SEC

HPMS SECTION

Roadlog

NUM

86

HPMS_SEG

HPMS SECTION SEGMENT

Roadlog

NUM

86

HPMS1

HPMS SECTION ID

Roadlog

NUM

86

HZM_IND

HAZARDOUS MATERIAL

Vehicle

CHA(1)

59

IMAG_NBR

IMAGE NUMBER

Accident

NUM

42

INJ

DRV/OCC INJURY

Occupant

CHA(8)

72

INSHTP1

INSIDE SHOULDER TYPE 1

Roadlog

CHA(1)

86

INSHTP2

INSIDE SHOULDER TYPE 2

Roadlog

CHA(1)

87

INSHWD1

INSIDE SHOULDER WIDTH 1

Roadlog

NUM

87

INSHWD2

INSIDE SHOULDER WIDTH 2

Roadlog

NUM

87

INT_NAME

INTERSECTING RTE NBR

Accident

NUM

42

INT_PREF

INTERSECT RTE PREFIX

Accident

NUM

42

INT_QUAD

INTERSECTION QUADRANT

Accident

NUM

43

INT_REL

INTERSECTION RELATED

Accident

CHA(1)

43

INT_TYPE

INTERSECTION FEATURE

Roadlog

CHA(1)

88

INTOX

ALCOHOL INVOLVED

Vehicle

CHA(1)

59

INV_DIR

INVENTORY DIRECTION

Roadlog

CHA(1)

88

KEY_RTE_APPRTE

KEY ROUTE APPURTENANCE NUMBER

Roadlog

NUM

88

KEY_RTE_APPURTCD

KEY ROUTE APPURTENANCE TYPE

Roadlog

CHA(1)

88

KEY_RTE_SEQNBR

KEY ROUTE SEQUENCE NUMBER

Roadlog

CHA(4)

88

KEY_RTE_STATION

KEY ROUTE STATION

Roadlog

NUM

89

KEY_RTE_SUF_CDE

KEY ROUTE SUFFIX CODE

Roadlog

CHA(1)

89

KEY_RTE_TYPCD

KEY ROUTE TYPE CODE

Roadlog

CHA(1)

89

LANEWID

AVERAGE LANE WIDTH

Roadlog

NUM

89

LIGHT

LIGHT CONDITION

Accident

NUM

43

LOC_TYPE

LOCATION TYPE

Accident

NUM

43

LPK_REST

PARKING RESTRICTIONS LEFT

Roadlog

CHA(1)

89

LST_SECD

LATEST CONSTRUCTION SECTION D

Roadlog

CHA(15)

89

LST_SECE

LATEST CONSTRUCTION SECTION E

Roadlog

CHA(10)

89

LST_UPDT

DATE OF LAST UPDATE

Roadlog

CHA(8)

90

MAIN_DIS

MAINTENANCE DISTRICT

Roadlog

NUM

90

MAIN_SEC

MAINTENANCE SECTION

Roadlog

CHA(6)

90

MAINTENC

MAINTENANCE RESPONSIBILITY[SS3] 

Roadlog

NUM

90

MED_TYPE

MEDIAN TYPE

Roadlog

NUM

90

 


SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

MEDWID

MEDIAN WIDTH

Roadlog

NUM

90

MILEPOST

MILE STATION

Accident

NUM

44

MISCACT1

DRV MISC ACTN 1 CD

Vehicle

NUM

59

MOSTHARM

VEHICLE MOST HARMFUL INVOLVEMENT

Vehicle

CHA(1)

61

MRK_BEG

MARKED BEGINNING

Roadlog

NUM

91

MRK_RTE1

MARKED ROUTE1

Roadlog

CHA(6)

91

MRK_RTE2

MARKED ROUTE2

Roadlog

CHA(6)

91

MRK_RTE3

MARKED ROUTE3

Roadlog

CHA(5)

91

MRK_RTE4

MARKED ROUTE4

Roadlog

CHA(5)

91

MRK_RTNBR

MARKED ROUTE NUMBER

Roadlog

NUM

91

MRKD_RTE_BEGMP

MARKED ROUTE BEGINNING MILEPOST

Roadlog

NUM

91

MULTICNT

AVERAGE ANNUAL DAILY MULTI-UNIT VOLUME

Roadlog

NUM

92

MUNI_NAME

MUNICIPAL NAME

Roadlog

CHA(20)

92

MVMT

MILLION VEHICLE MILES OF TRAVEL

Accident

NUM

44

MVMT

MILLION VEHICLE MILES OF TRAVEL

Roadlog

NUM

92

NAT_HWY

NATIONAL HIGHWAY SYSTEM

Accident

CHA(1)

44

NEW_ONEWAY

NEW ONEWAY INDICATOR

Roadlog

CHA(1)

92

NHS_CDE

NATIONAL HIGHWAY SYSTEM

Roadlog

CHA(1)

92

NO_LANES

TOTAL NUMBER OF LANES

Roadlog

NUM

92

NO_SPLNS

NUMBER OF SPECIAL LANES

Roadlog

NUM

93

NON_ATTN

NON-ATTAINMENT AREA

Roadlog

NUM

93

NUM_K

TOTAL KILLED IN VEHICLE

Vehicle

NUM

61

NUM_OCC

NO. OF OCCUPANTS IN VEHICLE

Vehicle

NUM

61

NUMINJ

TOTAL NUMBER INJURED IN VEHICLE

Vehicle

NUM

61

NUMVEHS

TOT-NBR-VEHICLES

Accident

NUM

44

OCC_AIR

OCCUPANT AIR BAG

Occupant

CHA(1)

73

OCC_IMAG

IMAGE NUMBER

Occupant

NUM

73

OCC_REEL

REEL NUMBER

Occupant

NUM

73

ODM_MILE

ODOMETER MILE

Roadlog

NUM

93

ODM_SIGN

ODOMETER SIGN

Roadlog

CHA(1)

93

OLD_AADT

OLD AADT

Roadlog

NUM

93

OLD_DATE

DATE

Accident

NUM

44

 

SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

ONEWAY

ONEWAY INDICATOR

Roadlog

NUM

94

OP_ID

OPERATOR ID

Accident

NUM

45

OPCRSNBR

OPPOSITE ROAD CRS NUMBER

Roadlog

NUM

94

OPP_FAULT

OPPOSITE ROAD FAULT

Roadlog

CHA(3)

94

OPP_PAVDIS

OPPOSITE ROAD PAVEMENT DISTRESS

Roadlog

CHA(10)

94

OPP_RUTDEPT

OPPOSITE ROAD RUT DEPTH

Roadlog

CHA(3)

94

ORG_SECB

ORIGINAL CONSTRUCTION SEC B

Roadlog

CHA(15)

94

ORG_SECC

ORIGINAL CONSTRUCTION SEC C

Roadlog

CHA(5)

94

OUTSHTP1

OUTSIDE SHOULDER TYPE 1

Roadlog

CHA(1)

94

OUTSHTP2

OUTSIDE SHOULDER TYPE 2

Roadlog

CHA(1)

95

OUTSHWD1

OUTSIDE SHOULDER WIDTH 1

Roadlog

NUM

95

OUTSHWD2

OUTSIDE SHOULDER WIDTH 2

Roadlog

NUM

96

OVHOBSNR

OVERHEAD OBSTRUCTION NUMBER

Roadlog

CHA(7)

96

PAV_DIST

PAVEMENT DISTRESS

Roadlog

CHA(10)

96

PAVECOND

PRESENT SERVICE RATING

Roadlog

NUM

96

PCNT_TRK

PERCENTAGE TRUCKS

Roadlog

CHA(2)

96

PED_AGE

AGE OF THE PED/PEDALCYCLIST

Vehicle

CHA(2)

61

PED_CLT

PED TYPE OF CLOTHING

Vehicle

CHA(1)

62

PED_FLAG

PEDESTRIAN FLAG

Vehicle

CHA(1)

62

PED_LOC

PED/PEDAL LOCATION

Vehicle

CHA(1)

62

PED_OTH

PEDESTRIAN/OTHER

Vehicle

NUM

62

PED_VIS

DRIVER VISION

Vehicle

CHA(1)

62

PEDACT

PED/PEDALCYCLIST ACTION/MOVEMENT

Vehicle

NUM

63

PERSON_TYP

PERSON TYPE

Vehicle

NUM

63

PHYSCOND

DRIVER PHYSICAL CONDITION

Vehicle

NUM

63

PLN_SEQ

PLANNING SEQUENCE

Roadlog

NUM

97

POP_GRP

MUNICIPALITY POPULATION GROUP

Roadlog

CHA(1)

97

POP_GRP

POPULATION GROUP

Accident

NUM

45

PRKLN_WD

PARKING LANE WIDTH

Roadlog

NUM

97

PTCONT1

POINT OF CONTACT #1

Vehicle

CHA(2)

64

RATE_DTE

MONTH-YR OF CONDITION RATING

Roadlog

NUM

97

RD_DEF

ROAD DEFECTS

Accident

NUM

45

RD_DIST

TOWNSHIP/ROAD DISTRICT

Roadlog

NUM

97

RD_STRUC

STRUCTURE NUMBER

Roadlog

CHA(7)

98

RD_YEAR

YEAR ROAD CONSTRUCTED

Roadlog

NUM

98

SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

RDSURF

ROAD SURFACE

Accident

NUM

45

REEL_NBR

REEL NUMBER

Accident

NUM

46

REF_PNT

REFERENCE POINT

Roadlog

CHA(15)

98

REF_PNT1

REFERENCE POINT 1

Roadlog

CHA(20)

98

REF_PNT2

REFERENCE POINT 2

Roadlog

CHA(20)

98

REF_PNT3

REFERENCE POINT 3

Roadlog

CHA(20)

98

REF_PNTA

REFERENCE POINT A

Roadlog

CHA(15)

98

REFPNT1A

REFERENCE POINT 1 TYPE

Roadlog

CHA(2)

98

REFPNT2A

REFERENCE POINT 2 TYPE

Roadlog

CHA(2)

98

REFPNT3A

REFERENCE POINT 3 TYPE

Roadlog

CHA(2)

98

REFPT1IN

REFERENCE POINT 1 INTERSECTION

Roadlog

CHA(1)

99

REFPT2IN

REFERENCE POINT 2 INTERSECTION

Roadlog

CHA(1)

99

REFPT3IN

REFERENCE POINT 3 INTERSECTION

Roadlog

CHA(1)

99

REPORT

REPORTABLE ACCIDENT

Vehicle

CHA(1)

64

RESEV_RD

RESERVATION ROAD

Roadlog

CHA(1)

99

RESIDLOC

RESIDENCE OF DRIVER

Vehicle

NUM

64

REST1

SAFETY EQUIPMENT

Occupant

NUM

73

REV_CDE

FORWARD/REVERSE CODE

Roadlog

CHA(1)

99

RODWYCLS

ROADWAY CLASSIFICATION

Accident

CHA(2)

46

RODWYCLS

ROADWAY CLASSIFICATION

Roadlog

CHA(2)

99

ROW

RIGHT OF WAY

Roadlog

NUM

99

RPK_REST

PARKING RESTRICTION

Roadlog

CHA(1)

100

RR_CRX

RAILROAD CROSS RIDEABILITY

Roadlog

NUM

100

RRD_LNK

RAILROAD LINK NUMBER

Roadlog

CHA(7)

100

RRX_ALP

RR CROSSING ALPA NBR

Accident

CHA(1)

46

RRX_DIRCD

RAILROAD DIRECTION CODE

Roadlog

CHA(1)

100

RRX_NBR

RAILROAD CROSSING NUMBER

Accident

NUM

46

RRX_RIDE

RAILROAD CROSSING RIDEABILITY

Roadlog

CHA(1)

100

RTE_APPURT

ROUTE APPURTENANCE

Roadlog

NUM

101

RTE_NBR

ROUTE NUMBER

Accident

NUM

46

RTE_NBR

ROUTE NUMBER

Roadlog

NUM

101

RTE_PREF

ROUTE PREFIX

Accident

NUM

46

RTE_SEGCD

ROUTE SEQUENCE NUMBER

Roadlog

CHA(2)

101

RTE_STAT

ROUTE STATION

Roadlog

NUM

101

RTE_STAT_END

ROUTE STATION END

Roadlog

NUM

101

RTE_SUFX

ROUTE SUFFIX

Roadlog

CHA(1)

101

RTE_TYPE

ROUTE TYPE

Accident

NUM

47

RTE_TYPE

ROUTE TYPE

Roadlog

NUM

101

SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

RURURB

RURAL / URBAN CODE

Roadlog

NUM

102

RUT_DEPTIN

RUT DEPTH INDICATOR

Roadlog

CHA(3)

102

S_INVLOC

SECOND INVOLVEMENT LOCATION

Vehicle

NUM

65

S_RTENBR

SAF-MRK-ROUTE NUMBER

Roadlog

NUM

102

S_RTETYP

SAF-MRK-ROUTE TYP

Roadlog

NUM

102

SAF_CNTL

SAF-ACCESS-CNTL

Roadlog

NUM

102

SAF_FASY

SAF-FASYS

Roadlog

NUM

103

SAF_TWN

SAF-TWNSHP

Roadlog

NUM

103

SEATPOS

SEATING POSITION

Occupant

NUM

74

SEG_LNG

SEGMENT LENGTH

Deficient Curve

NUM

113

SEG_LNG

SEGMENT LENGTH

Roadlog

NUM

103

SEV_CDE

SEVERITY CODE

Accident

NUM

47

SEVERITY

COLLISION SEVERITY

Accident

NUM

47

SEX

OCCUPANT SEX

Occupant

NUM

74

SHLD_CON

SHOULDER CONDITION

Roadlog

NUM

103

SND_INV

SECOND INVOLVEMENT

Vehicle

NUM

65

SOB_TEST

FIRST SOBRIETY/CONDITION

Vehicle

NUM

65

SPD_LIM2

POSTED SPEED LIMIT (MINUS DIRECTION)

Roadlog

NUM

103

SPD_LIMT

ROADWAY SPEED LIMIT

Roadlog

NUM

103

SPEC_SYSM

SPECIAL SYSTEMS

Roadlog

CHA(1)

104

SPEC_VEH

SPECIAL VEHICLE

Vehicle

NUM

65

SPLN_TYP

LANES SPECIAL TYPE

Roadlog

CHA(1)

104

SPLN_WID

LANES SPECIAL WIDTH

Roadlog

NUM

104

STAT_DIR

PLAN STATION DIRECTION

Roadlog

CHA(1)

104

STOU_IND

STRUCTURE OVER/UNDER IND

Roadlog

CHA(1)

105

STR_ENDM

STRUCTURE END MILEPOST

Roadlog

CHA(5)

105

STR_LNG

STRUCTURE LENGTH

Roadlog

NUM

105

STRDIRCD

STRUCTURE DIRECTION CODE

Roadlog

CHAR(1)

105

STRK_CDE

STRIKE STRUCK CODE

Vehicle

CHA(1)

66

STRT_NAM

STREET-NAME

Roadlog

CHA(15)

105

STRU_FAC

STRUCTURE FACILITY LOCATION

Roadlog

CHA(20)

105

STRU_LNK

STRUCTURE LINK NUMBER

Roadlog

NUM

105

SUF_CDE

SUFFIX CODE

Roadlog

CHA(1)

105

SURF_RAT

SURFACE CONDITION RATING

Roadlog

NUM

106

SURF_TYP

SURFACE TYPE - ROAD 1

Roadlog

NUM

106

SURF_WID

TOTAL SURFACE WIDTH

Roadlog

NUM

106

SURF_YR

YEAR OF PRESENT SUF CONST

Roadlog

CHA(2)

107

SAS VARIABLE NAME

DESCRIPTION

SAS VARIABLE FILE

FORMAT TYPE

PAGE NO.

SURFDATE

YEAR OF PRESENT SUF CONST

Roadlog

CHA(4)

107

T_INVLOC

THIRD INVOLVEMENT LOCATION

Vehicle

NUM

66

TC_COND

TRAFFIC CONTROL CONDITION

Accident