Scottish Government

Contributory Factors

Table J: Contributory factors ( CFs), 2007

( a ) Overall summary

( b ) Most often-recorded CFs - all road users (from Table K)

( c ) Most often-recorded CFs - main types of road user (from Table L)

( d ) Most often-recorded CFs - by the severity of the accident (#)(from Table N)

(*) most of these numbers do not appear in the other tables
(D/R) indicates Driver / Rider (to distinguish this Contributory Factor from a similar one which applies to Pedestrians)
( $ ) including some pedestrian only Contributory Factors that were allocated to a vehicle
( # ) excluding any repeats of the same Contributory Factor within an accident - e.g. if two of the participants in an accident were thought to be exceeding speed limit, that Contributory Factor is only counted once for the purpose of producing this table

CONTRIBUTORY FACTORS

• Section 1: Introduction
• Section 2: Collection of CF data
• Section 3: CFs by type of participant;
• Section 4: CFs of accidents
• Section 5: CFs of killed and serious injuries - by road user type

1. Introduction

1.1 The Stats 19 road accident statistics system has always collected a range of information about the circumstances in which accidents occur, the vehicles that were involved, their manoeuvres, and the resulting casualties. However, such data primarily answer questions about the where, when and who of accidents. Insight into the why and how may be provided by the Contributory Factor ( CF) data, which were added to the Stats 19 system in 2005 in the biggest single change ever made to the Stats 19 collection.

1.2 The aim of collecting the CFs is to help to identify the key actions, failures and occurrences that led directly to the actual impact: the data may show why the accident occurred, and may give clues about how it might have been prevented (but this cannot always be the case). The factors recorded in the Stats 19 returns:

• reflect the reporting officer's opinion at the time of reporting the accident (or the opinion of a person whose duties include deciding which CFs should be recorded based on the officer's report).
• are based on the information which was available at that time, so may not be the result of subsequent extensive investigation (indeed, subsequent enquiries could result in the reporting officer's opinion changing).

Some CFs may be less likely than others to be recorded, since clear evidence of them may not be available, or may be very difficult to obtain, after an accident has occurred (e.g. in the case of the nervous, uncertain or panic factor). Participants and witnesses may provide incomplete or conflicting accounts of what happened. The CF data therefore depend upon the skill and experience of the reporting officer to reconstruct the events which led directly to the accident, and so are more subjective in nature than other Stats 19 data. This should be kept in mind when using these results.

1.4 The main points shown by the CF data for 2007 include:

• a typical accident does not have a single cause - on average, there were more than two CFs per accident (although this varies between Police Forces - see paragraph 2.5);
• on average, more CFs were recorded for fatal accidents and fewest for slight accidents;
• the most frequently-used CFs were:
  • (driver/rider) failed to look properly;
  • loss of control;
  • (driver/rider) failed to judge other person's path/speed;
  • (driver/rider) careless / reckless / in a hurry;
  • slippery road (due to weather);
  • poor turn or manoeuvre;
  • travelling too fast for the conditions;
  • sudden braking
  • following too close and
  • pedestrian careless/reckless/in hurry
• the CFs most often associated with fatal accidents were:
  • loss of control;
  • travelling too fast for the conditions
  • (driver/rider) failed to look properly;
  • (driver/rider) careless / reckless / in a hurry
  • poor turn or manoeuvre
  • exceeding speed limit;
  • pedestrian failed to look properly ;
  • (driver/rider) failed to judge other person's path/speed ;
  • inexperienced or learner driver/rider;
  • (driver/rider) impaired by alcohol and
  • (pedestrian) impaired by alcohol

More detailed information about the results appears in subsequent sections of this note.

2. The collection of the Contributory Factor data

2.1 The Instructions for the Completion of Road Accident Reports (page 10 & pages 84 - 101 the Stats 20 document) include the following points on CFs:

• the CFs reflect the reporting officer's opinion at the time of reporting, and may not be the result of extensive investigation;
• it is recognised that subsequent enquiries could result in a change in the reporting officer's opinion;
• the CFs are largely subjective, and depend upon the skill and experience of the investigating officer to reconstruct the events which led directly to the accident;
• the need to exercise judgement when recording CFs is unavoidable;
• CFs should be identified on the basis of evidence from sources such as witness statements and vehicle and site inspections;
• the evidence may be of variable quality, so the officer should record very likely or possible for each CF;
• when there is conflicting evidence (e.g. conflicting witness statements), the reporting officer should decide on the most credible account of the accident and base the codes on this, taking into account all other available evidence.

2.2 Regardless of the number of vehicles that were involved in the accident, at most six sets of CF data can be recorded per accident. Each set contains three pieces of information:

• a factor which is thought to have contributed to the occurrence of the accident - selected from list of 77 , such as:
  • exceeding speed limit ( CF code 306);
  • travelling too fast for the conditions (307);
  • failed to look properly (405);
  • impaired by alcohol (501);
  • impaired by drugs (illicit or medicinal) (502)
• the participant in the accident to whom the factor is related:
  • whether this is a:
    • Vehicle - in which case the factor may relate to the driver/rider or to the road environment;
    • Casualty - a pedestrian or a passenger in a vehicle; or
    • Uninjured pedestrian.
  • if a Vehicle or a Casualty, the relevant Stats 19 reference
• whether it was thought very likely or just possible that this factor contributed to the occurrence of the accident

As the example below indicates:

• more than one factor may be recorded for the same participant; and
• any given factor may be recorded for two or more different participants;

subject only to the limit of a maximum of six sets of CF data per accident

2.3 The illustrative Stats 19 form that appears in Annex B includes a page that shows all the CF codes and their descriptions, along with a brief set of instructions for the completion of the CF part of the Stats 19 returns. More detailed information is available in Stats 20: pages 84 to 101 contain detailed information about the individual CF codes, and page 10 sets out basis on which they should be recorded (some points from which were mentioned above, in paragraph 2.1) and the procedure for allocating them - for example:

• the CFs may be recorded in any order (so nothing can be inferred from the order in which they appear);
• more than one CF may be related to the same road user; and
• the same CF may be related to more than one road user.

2.4 Clearly, there could be a lot of CF information in the case of an accident which involved several vehicles, if it was thought that several of them contributed to its occurrence. The following is an example of the potential complexity of the CF data. Car 1 is rapidly travelling along a straight road when Car 2 suddenly appears in front of it, having emerged from a pub car park. The driver of Car 1 brakes sharply, to avoid a collision. As Car 2 drives off, Car 1 is hit from behind by a motorcycle, whose rider and passenger are both killed. The following might be recorded as the CF data for this accident:

This accident has three participants and sixCFs, two of which are the same (exceeding speed limit) but apply to different participants (Car 1 and Motorcycle). This example will be referred to from time to time, when describing some of the CF results.

2.5 Because the CFs were added to the Stats 19 data specification at the start of 2005, the results for 2005 could have been affected by teething troubles. In June 2006, the Liaison Group on Road Accident Statistics ( LGRAS) discussed a paper on aspects of the quality of the data, some points from which were mentioned in the corresponding article in an earlier of this publication. It also remains the case that the extent to which CFs are used seems to vary between Police Forces. In 2007, excluding the few accidents for which there were no CFs, there were, for example 2.1 CFs (on average) per accident for Scotland as a whole - this varied between 1.5 and 2.6 between Forces. In addition, while most Police Forces' CFs are allocated by the reporting officer, in one Force they are allocated by a small team of specialist crash investigators, who read the officers' reports and then allocated the CFs. It was suggested that the data might be more accurate for fatal and serious accidents than for slight accidents, because the former would generally be attended by more experienced road policing officers.

2.6 The meeting also discussed the fact that there was inconsistency between the CF code and the Type of Participant code in (at that time) about 3-4% of cases. The most frequent problem was the combination of the CF code for pedestrian failed to look properly with the Type of Participant code for a Vehicle. In general, in such cases, one cannot deduce from the data whether it is the CF code that is wrong (i.e. the code for driver/rider failed to look should have been used) or the Type of Participant code that is wrong (i.e. the code for a pedestrian Casualty or an Uninjured pedestrian should have been used). The then Scottish Executive therefore introduced additional computer cross-checks of the data submitted by Police Forces, in order that such cases would be queried as and when they arose, which led to an improvement in the quality of this aspect of the CF data (there appears to be inconsistencies in only about 1% of cases in the data for 2007), and should produce further improvements in future years.

2.7 There may be other changes in some of the patterns of the reporting of CFs, as a result of such discussions, the introduction of additional computer cross-checks of the data, Police Forces' increasing experience of the collection and recording of such information, and the use of the data by the Police, local authorities and central government.

3. Numbers of Contributory Factors reported (Tables J (a)-(c) & Tables K-M)

3.1 The basis of the figures

3.1.1 Data relating to all reported CFs were used to produce Tables J (a) to (c), and Tables K to M. In cases where the same CF applies to more than one vehicle in the same accident, it is counted once for each of them. As a result, the example described in paragraph 2.4 would be counted six times in Tables J (a) to (c) and in Tables K and L. For example, in Table L, it would be counted:

The example does not have any CFs associated with a Casualty or with an uninjured pedestrian - but, if it did, they would be counted in the Other persons column in Table L.

3.1.2 The basis of these tables therefore differs from that of the later tables (those analysing the numbers of accidents and casualties, which were produced from data which exclude repeats of the same CF within an accident).

3.1.3 Table L shows that (as mentioned in paragraph 2.6) some pedestrian only CFs have been allocated to vehicles, and some vehicle CFs have been allocated to casualties or to uninjured pedestrians. As mentioned earlier inconsistent cases represent only a small percentage of the total, and it is hoped that they will be reduced further in future by the continued cross-checking of the codes for the CF and the Type of Participant.

3.1.4 Table M shows the combinations of CFs that were most often recorded for the same participant. A combination of two CFs (e.g. X and Y) could be reported in two ways - either as X+Y or as Y+X. The convention adopted for the purpose of producing this table is that the CF with the lower Stats 19 code value should appear as the firstCF in the combination, regardless of the order in which the CFs were reported. The example in paragraph 2.4 would produce three combinations of CFs, each of which could be counted in Table Three:

Following the convention adopted for the purposes of this table, the combination for Car 2 would be counted as failed to look properly + impaired by alcohol even though the two CFs appeared in a different order in paragraph 2.4.

3.2 Table J (a) to (c)

3.2.1 Table J (a) shows that 12,485 accidents were reported in 2007. CFs were recorded for 12,465 accidents, and 20 accidents did not have any CFs recorded. In total, 26,525 CFs were recorded, so there were more than twice as many CFs as accidents for which CFs were recorded. The average of 2.13 CFs per accident for which CFs were recorded provides a reminder that a typical accident is the result of the combination of two or more CFs, so does not have a single cause.

3.2.2 The remainder of Table J (a) shows the types of participant for whom the CFs were recorded. Over four-fifths (22,512) were vehicles (drivers/riders or the road environment), and about a seventh (3,511) were pedestrians who were casualties. Relatively few were passengers (162 or 0.6%) or uninjured pedestrians (232 or 0.9%).

3.2.3 Table J (b) lists the ten CFs which were recorded most often, overall. The first seven were:

Together, these seven CFs accounted for 39% of all the CFs that were recorded. None of the other CFs accounted for more than 5% of the total.

3.2.4 Table J (c) lists the most often-recorded CFs for each of the main types of road user. The total of the figures for a particular CF (e.g. failed to judge other person's path/speed) will be less than that given in part (b) for two reasons: first, some types of road user are not shown in part (c); and, second, some of the CFs do not appear among the most often-recorded for each of the types of road user. The figures show some differences between types of road user - for example, failed to look properly is the most often-recorded CF overall, and for most of the main types of road user, but loss of control is recorded most often for motorcyclists and sudden braking for bus drivers.

3.3 Contributory Factors listed in order of their frequency of use - Table K

3.3.1 Table K lists all 77 CFs in descending order of their frequency of use. The most often-used have already been described, as they appeared in Table J (b). Table K shows that some CFs are not used often - for example, many were used fewer than 100 times.

3.3.2 This table also shows how often each CF was described as very likely, and how often as possible. Overall, almost three-quarters of CFs (72%) were described as very likely, but the percentage varied markedly between different CFs. Leaving aside those which were used fewer than 100 times, the following were described as very likely on at least 86% of occasions on which they were used:

and the following were described as very likely on fewer than 60% of the occasions on which they were used:

3.4 Contributory Factors recorded - by type of participant - Table L

3.4.1 This table shows all the CFs, grouped in the way in which they appear in the illustrative Stats 19 form. The penultimate row gives the total number of vehicles of each type that were involved in all the accidents that were reported in the Stats 19 returns (including accidents for which no CFs were recorded, and vehicles to which no CFs were assigned), and the final row shows the average number of CFs reported per vehicle of each type. The most often-recorded CFs for the main types of road user have already been identified in Table J (c).

3.4.2 Table L shows that many of the CFs were recorded very rarely for most of the types of road user, and some marked differences between the CFs that were assigned to different types of vehicle - such as:

The final row shows that, on average, relatively few CFs were recorded per vehicle involved in injury accidents (including those without CFs) for pedal cycles (0.70) and the bus/coach/minibus grouping (0.72), and that this average was highest for motorcycles (1.21).

3.5 The combinations of Contributory Factors that were recorded most often for the same participant-Table M

3.5.1 This table lists the combinations of CFs for the same participant that were recorded at least 100 times, produced on the basis described in paragraph 3.1.4. The most frequently-occurring combination is (driver/rider) failed to look properly + (driver/rider) failed to judge other person's path/speed, which was recorded on 713 occasions. As would be expected, the CFs that were identified earlier as being used most often tend to appear in several of the most frequently-occurring combinations - for example, (driver/rider) failed to look properly and loss of control each occur in three of the ten most frequently-occurring combinations. However, the numbers indicate that even the most frequently-occurring combination of CFs arose in only a small percentage of all accidents.

4. Numbers of accidents in which each Contributory Factor was reported: Table J (d) and Table N

4.1 Table N shows the numbers of accidents to which each CF was thought to have contributed. For this purpose, each accident is counted at most once for each CF - even in cases where a particular CF applies to more than one of the participants in the accident. For example, the accident described in paragraph 2.4 is counted only once against exceeding speed limit, even though exceeding speed limit was recorded for two of the participants (counting the repeats of exceeding speed limit within that accident would exaggerate the number of accidents to which that CF contributed). Therefore, repeats of the same CF within an accident were excluded from the data used to prepare this table.

4.2 On this basis, the example accident would be counted five times:

4.3 As mentioned earlier, a typical accident (for which there is no obvious single cause) has more than one CF, and there are an average of over two CFs per accident which has any CFs recorded. As a result, the total of all the percentages in the column for accidents of a particular severity would be much greater than 100%. The final row of Table N shows that the average number of CFs per accident (excluding repeats) is highest for fatal accidents (2.45) and lowest for slight accidents (2.04).

4.4 It is incorrect to sum any two CFs to get the total number of accidents to which those factors were thought to have contributed, because there could be a number of accidents for which both factors were recorded (and would therefore double count).

4.5 Table J (d) lists the most often-used CFs for each category of accident severity. For fatal accidents, these were:

As there was an average of 2.45 such CFs per fatal accident, percentages will sum to over 100%.

4.6 Table J also provide the Cfs for other severities of accidents. The most common are:

These figures are slightly lower than those which appear in Table J (b) and Table K, due to the exclusion of any repeats of the same CF within an accident.

4.7 Table N shows how the incidence of some CFs varies with the severity of the accident. For example: loss of control is cited in 18% of all accidents for which CFs were recorded but 40% of fatal accidents; slippery road due to weather is cited in 12% of all accidents but 5% of fatal ones; travelling too fast for the conditions is cited in 10% of all accidents but 18% of fatal ones and exceeding speed limit is cited in 4% of all accidents but 11% of fatal ones.

5. Numbers of deaths and serious injuries resulting from accidents in which each Contributory Factor was reported: Tables O and P

5.1 Table O shows the numbers of deaths resulting from accidents in which each CF was reported. For this purpose:

• as with Table N, repeats of the same CF within an accident were excluded from this table - even in cases where a particular CF applies to more than one of the participants in that accident. E.g. the example accident (paragraph 2.4) would be counted only once against exceeding speed limit (even though exceeding speed limit was recorded for two of the participants) as the number of associated deaths would be exaggerated otherwise.
• every death should be counted against each of the CFs reported for the accident (excluding repeats), as each of those CFs was thought to have contributed to the occurrence of the accident causing those deaths

5.2 The accident which was described in paragraph 2.4 involved two deaths (the motorcycle rider and passenger). For the purpose of producing Table O, counting all the deaths against each of the CFs (apart from the repeats) means that:

• two deaths are counted against exceeding speed limit;
• two deaths are counted against impaired by alcohol;
• two deaths are counted against failed to look properly;
• two deaths are counted against sudden braking;
• two deaths are counted against following too close

So an accident in which two people died contributes ten to the grand total. The final row of Table O shows that the result is that the total of the numbers reported in the table is considerably greater than the actual number of deaths.

5.3 It is incorrect to sum the figures for any two CFs to get the total number of deaths to which those CFs were thought to have contributed. There may be a number of accidents for which bothCFs were recorded (and summing would result in double-counting)

5.4 Table O shows the CFs associated with the largest numbers of deaths were:

• loss of control - 121 deaths (representing 43% of all deaths in accidents for which CFs were recorded);
• travelling too fast for the conditions - 53 (19%);
• (driver/rider) failed to look properly - 50 (18%);
• (driver/rider) careless/reckless/in a hurry - 41 (15%);
• exceeding speed limit - 36 (13%);
• poor turn or manoeuvre - 36 deaths (13%)

5.5 Table P shows the CFs associated with the largest numbers of serious injured were

• loss of control - 565 serious injuries (representing 24% of all serious injuries in accidents for which CFs were recorded);
• (driver/rider) failed to look properly - 513 serious injuries (22%);
• (driver/rider) careless / reckless / in a hurry - 376 (16%);
• pedestrian failed to look properly - 354 (15%);
• travelling too fast for the conditions - 290 (12%)

Table K: All contributory factors recorded ( in order of frequency), 2007

(D/R) indicates "Driver / Rider" (to distinguish this Contributory Factor from a similar one which applies to Pedestrians)

Table L: All contributory factors recorded, by type of participant to which assigned, 20070

1. The "pedestrian only" CFs allocated to Vehicles, and the "vehicle only" CFs allocated to casualties or uninjured pedestrians, are presumed to be the result of errors in the data submitted by Police Forces
2. The total number of CFs may differ slightly from that in Table One, due to the exclusion of a small number of cases for which there was an invalid value of (e.g.) the code for the type of participant or (for a Vehicle) the vehicle reference number.

Table M: Combinations of CFs recorded most often for the same participant, 2007

NB: the basis upon which the combinations are produced is described in the text.
However, an additional example may be helpful.
Suppose that the "defective brakes" CF has been allocated to one participant (A, say)
the "failed to look properly" CF has been allocated to two participants (A and B, say) and
the "failed to judge other person's path/speed" CF has been allocated to three participants (A, B and C, say)

The following combinations of CFs would be allocated to the same participant:
A defective brakes + A failed to look …
A defective brakes + A failed to judge …
A failed to look ... + A failed to judge …
B failed to look ... + B failed to judge …

Table N: Numbers of accidents for which each CF was recorded, 2007

Table O: Numbers of deaths - in accidents for which each CF was recorded, 2007

Table P: Number of serious injuries - in accidents for which each CF recorded, 2007