DRC creates a number of output files, depending on the command line options specified. A directory will be created to contain these output files, the name of which will usually be the concatenation of either the word being simulated, or the name of the batch file, and the extension ".drc". For example, a simulation of the single word "kyoot", will result in a directory called "kyoot.drc". A batch job executed from a batch file named "mybatch" will result in a directory named "mybatch.drc". If a directory with the same name already exists, DRC will append "-1" or "-2" etc. to the directory name until it finds a name for a directory that does not exist.

Inside the ".drc" directory, there will be at least two files: a "parameters" file, and a ".RT" file. If the "-a" command-line option was used, there will also be one ".acts" file for each word simulated. The parameters file is a text file containing the parameters used during the simulation. The format of the other files is outlined below.

The .RT file contains a preamble, specifying a few important pieces of information about the simulations run. Each line in the preamble is preceded by a semi-colon. Here is an example:

; DRC Reaction Time Results File
; Simulation started: Thu Mar 24 14:44:46 2005
; DRC Model version 1.1.0
; - Visual Feature and Letter Layer units: 8 Phoneme Layer: 8
; Language: English1.1
; - Letters: 27 Visual Features: 14 Phonemes: 45 Stress Types: 0
; - Orthographic Lexicon entries: 7985 (skipped 10 that were too large)
; - Phonological Lexicon entries: 7127 (using COMBINED entries for homophones)
; - Multi-Letter rules: 65 Context-Sensitive rules: 17 Two-Letter rules: 95
; - One-To-Multi rules: 2 Single-Letter rules: 38 Output rules: 8
; - Body rules: 829 Stress rules: 0
; Results for batch "ExampleBatchFile"
    

Following this preamble there will be one line for each word simulated. Each line has 5 or 6 fields. They are:

1. The word simulated
2. The final pronunciation produced by DRC
3. The number of cycles for which the word was simulated
4. Either NAMED or LOWAC (see below)
5. Either CORRECT or WRONG (see below)
6. The amount of time the simulation took in milliseconds (if the -t argument is used)
7. In batch jobs, any comments for a word in the batch file are repeated here

For field 3, LOWAC indicates that activation levels in the phoneme layer did not reach the critical levels needed for the input to be considered "named". The critical levels required are defined by the MinNamingPhonology parameter.

Field 5 contains an indicator of whether the final pronunciation produced by DRC was correct. If the input was a word in DRC's vocabulary, CORRECT means that the pronunciation produced by DRC matched one of the pronunciations for that word in the vocabulary. If the input was a nonword, CORRECT means that the final pronunciation produced by DRC matches the pronunciation produced by DRC's GPC rules for the nonword.

If the extra output option (-e) is used, the output will contain the following fields.

1. The word simulated
2. An indicator of whether the input is a WORD, NONWORD or a HETEROPHONIC word
3. The final pronunciation produced by DRC
4. An indicator of whether DRC's output is a phonological word (PWORD) or nonword (PNONWORD)
5. The number of cycles for which the word was simulated
6. Either NAMED or LOWAC (see above)
7. Either CORRECT or WRONG (see above)
8. The correct pronunciation
9. The number of correct phonemes in DRC's output
10. The number of phonemes in the correct pronunciation
11. The number of incorrect phonemes in DRC's output
12. The amount of time the simulation took in milliseconds (if the -t argument is used)
13. In batch jobs, any comments for a word in the batch file are repeated here

In field 2, HETEROPHONIC indicates that the input was a word in DRC's vocabulary which is a heterophonic homograph.

The correct pronunciation given in field 8 will be the regular (GPC rule based) pronunciation if the input was a nonword. If the input was a word then it will be either the pronunciation for the input that matches DRC's output, if DRC's output was correct, or the pronunciation with the most matching phonemes, with higher frequency pronunciations selected in the case of a tie.

These files contain information about the levels of activation that occured during the simulation. There will be a preamble, similar to the preamble for the .RT file, at the start of each .acts file. Following that will be many lines indicating activation levels at different stages of the simulation. Only activation levels greater than or equal to the minimum are reported.

At the end of each cycle, the letter layer, orthographic lexicon, phonological lexicon, GPC route and phoneme layer will have activation levels reported in order. After this a set of totals for each layer will be reported.

For the letter layer, GPC route, and phoneme layer, there will be one line for each letter or phoneme with a sufficiently high activation. Each line will contain 4 fields:

1. The cycle
2. A letter, GPC output phoneme or phoneme position indicator (e.g.s L0 = letter position 0, GPC3 = GPC route output phoneme position 3, P4 = phoneme layer position 4)
3. The activation level
4. The name of the letter or phoneme

For example:

Cycle140 L2 0.999600 A

Indicates that at cycle 140, the letter 'A' at position 2 in the letter layer had an activation level of 0.999600.

For the two lexicons, there will be a line for each lexical unit with sufficiently high activation. Each line will contain 4 fields:

1. The cycle
2. Either "Orth" or "Phon" depending on which lexicon the unit is in
3. The activation level
4. The name of the unit

Examples:

Cycle140 Orth 0.016596 THAW
Cycle140 Phon 0.793794 D{t

Indicate that at cycle 140, the word THAW in the orthographic lexicon had an activation of 0.016596, while the phonology /D{t/ in the phonological lexicon had an activation of 0.793794.

Prior to the reporting of the output activation level of the GPC route, a single line will appear showing which GPC rules are being used by the GPC route. This line has 4 fields:

1. The cycle
2. The string "GPCRoute"
3. The input to the GPC route
4. A list of the rules being applied

For example:

Cycle140 GPCRoute THAT++++ TH->T A->{ T->t

Inidicates that at cycle 140, the input to the GPC route was THAT++++ (+ is the 'blank' indicator) and matching rules were: Letters TH produce output phoneme /T/, the letter A produces output phoneme /{/ and the letter T produces output phoneme /t/.

The totals are reported one on each line. There are 3 fields on the totals lines:

1. A T followed by a layer indicator (L for letter layer, O for orthographic lexicon, P for phonological lexicon, GPC for the GPC route, and Ph for the phoneme layer)
2. The cycle
3. The total activation in the layer

Examples:

TL Cycle140 8.735193
TO Cycle140 1.046514
TP Cycle140 0.793794
TGPC Cycle140 6.868971
TPh Cycle140 5.274763

Inidcate that at cycle 140 the letter layer had a total activation of 8.735193, the Orthographic Lexicon has a total activation of 1.046514, the Phonological lexicon had a total activation of 0.793794, the total output activation of the GPC route was 6.868971 and the total activation in the phoneme layer was 5.274763.