SL-TRC-Data

The data from the Sierra Leone Truth and Reconciliation Commission is contained in a comma-delimited ASCII file. The tables below describe the variables in the data file.  Note that to protect the privacy of the people who suffered, all personal identifying information about the victims or witnesses of the violations described in the data has been removed.

The unit of analysis for each record in this structure is the violation.

Each violation was of a particular type, happened at a particular time and place, and was committed by zero, one, or several organizational perpetrators. The violation was committed against zero or one named (individually identified) victims, and zero or more anonymous (unidentified) additional victims.

In order to count the number of violations recorded in the dataset, users should simply count records. Each violation may have been committed against one or many persons in the same place at the same time. In the cases where one violation was committed against mare than one person, the records share the same actid. The variable actorid identifies unique victims, so in order to count victims, the user should count the unique values of actorid. Note that the value of actorid is repeated for the cases where a victim suffered more than one violation.

Perpetrator Aggregations

This is the perpetrator classification list used by the TRC to classify perpetrators cited in witness testimonies. Each of the P* variables contains a 1 if this perpetrator was involved in this violation; otherwise 0 is recorded.

Pd_* list all the actors in the Sierra Leonan conflict (see the table below).

P3_* is a refined aggregation separating the AFRC from the SLA. In this case SLA means the Sierra Leonean national army prior to the 25^th May 1997 coup. The AFRC refers to the post-coup soldiers who initially formed a government and then fled Freetown after the ECOMOG intervention. Note that the Westside boys are considered to be aligned with the AFRC, not the RUF.

To complete the aggregation, the following operations are applied to the data:

• AFRC abuses prior to the 25^th May 1997 are moved to the SLA category

• SLA abuses from the 25^th May 1997 are moved to the AFRC category

• Ambiguous abuses occurring at some unknown time in 1997 are moved to an additional category AFRC/SLA as the correct group is ambiguous.

Arguably the AFRC/SLA category could be eliminated, with the abuses categorized as AFRC. The abuse profile for the AFRC/SLA category is comparable to that of the AFRC.

The 'unknown' category covers abuses where no mention was made of the responsible institution or cases where the presence of several perpetrator groups make it impossible to distinguish which was responsible for a specific abuse.

Perpetrator Aggregation Table

Ethnicity Aggregations

This is the ethnicity classification list used by the TRC to classify the ethnicity of the statement-giver and of the people he or she cited in the statement to the TRC.

* ethn1: The initial aggregation of ethnicity was designed to test suggestions that groups have targeted victims according to a broad allegiance, either to the North or South; in other words, Temne and Mende respectively. The proposed ethnic allegiances were verified by researchers.

ethn2t: The second aggregation was used to consider more specific targeting according to ethnicity. For example, other sources suggested that the NPFL targeted the trading peoples – Madingo and Fullah – because of the perception (by perpetrators) that they were loyal to Samuel Doe.

The location codes used in the data file are contained here.

For the Sierra Leone TRC's evaluation of this data, see Statistical Appendix (25 March 2005) Adobe Acrobat PDF format, 407 Kb