I. Introduction


During the California murder case of People v. O.J. Simpson, DNA results indicated that blood found on a rear gate at the crime scene belonged to O.J. Simpson and that blood found on socks at Simpson's residence belonged to murder victim Nicole Brown Simpson. To counter this evidence, the defense maintained that the police had planted this blood, using blood samples taken from O.J. Simpson and the body of Nicole Brown Simpson after the murders. Authorities collected those blood samples in test tubes that contained a blood preservative, ethylenediamine tetraacetic acid (EDTA). Accordingly, to disprove the defense's assertion, the prosecution asked the FBI to determine whether the blood evidence from the rear gate and socks contained levels of EDTA consistent with preserved blood from these test tubes.


CTU Chief Roger Martz and several research chemists at the FBI Forensic Science Research Unit (FSRU) in Quantico, Virginia, worked to develop a method for identifying EDTA in blood. Martz later examined the bloodstain evidence from the rear gate and socks and concluded that they did not contain EDTA-preserved blood. The defense subsequently called Martz to testify about this work.


Whitehurst stated that after Martz testified, scientists at the FSRU were highly critical of several aspects of Martz's testimony. These scientists reportedly claimed that Martz committed perjury by testifying that he had developed the method used to examine the evidence, misled the jury concerning the FSRU's validation studies and events surrounding the development of the protocol, misled the defense by stating that all digital data from the analysis of the evidence had been erased, and generally testified in an arrogant manner.


In connection with these allegations, we obtained and reviewed transcripts of Martz's trial testimony and Whitehurst's testimony during a hearing in the Simpson case. We also obtained and reviewed a videotape of part of Martz's testimony. We further reviewed pertinent Laboratory reports, dictation, work papers, and charts in the case, along with various correspondence and memoranda. Finally, we interviewed numerous personnel in the FSRU and Laboratory, including former Laboratory Director Milton Ahlerich, Acting Laboratory Director Donald Thompson, former SAS Chief Kenneth Nimmich, SAS Section Chief Randall Murch, FSRU Chief Bruce Budowle, FSRU Quality Assurance Chief Larry Presley, CTU Chief Roger Martz, research chemists Dr. Mark Miller, Dr. Dean Fetterolf, Dr. Bruce McCord, Dr. Mary Tungol, and Whitehurst.


We find no basis to conclude that Martz committed perjury or misled the jury or defense in the Simpson case. Nor do we conclude that Martz improperly erased digital data from his results. We do conclude that because of his lack of preparation, his deficient record-keeping and note-taking practices, and certain aspects of his presentation and demeanor at trial, Martz poorly represented the Laboratory and the FBI in this case.


II. Factual Background


The prosecution's DNA test results in the Simpson case indicated that O.J. Simpson's blood was found on a rear gate at the murder scene and that Nicole Brown Simpson's blood was found on socks taken by police from O.J. Simpson's residence. The defense suggested in their opening statement that the police had planted this evidence, using known blood samples from Nicole Brown Simpson and O.J. Simpson. The police had collected these known blood samples in test tubes containing the preservative EDTA.


To refute the defense's contention, in late 1994 one of the Simpson prosecutors asked Dr. Bruce Budowle in the FSRU to determine whether the bloodstain evidence from the gate and socks contained the preservative EDTA. In January 1995, at Budowle's direction, various research chemists within the FSRU began working to develop a method for identifying EDTA-preserved blood. Shortly thereafter, Martz also began working on the same task at the Laboratory in Washington, D.C. By early February 1995, both the chemists in the FSRU and Martz had developed methods, using slightly different procedures, for isolating and identifying EDTA in blood samples.


In mid-February, 1995, the prosecution sent Martz various items for testing, including a cloth swatch with blood from the rear gate (Q204); a pair of socks with blood taken from O.J. Simpson's residence (Q205/206); a known blood sample from Nicole Brown Simpson (K67); and a known blood sample from O.J. Simpson (K68). Martz examined blood from these items using liquid chromatography/tandem mass spectrometry (LC/MS/MS) and high pressure liquid chromatography (HPLC). In his dictation, Martz reported that EDTA was present in the preserved blood samples taken from Nicole Brown Simpson and O.J. Simpson, but not in the bloodstain evidence from the rear gate and socks, in pertinent part, as follows:


EDTA, a blood preservative, was identified on stains prepared from the K67 and K68 blood samples from Nicole Simpson and Orenthal J. Simpson, respectively. No EDTA was identified in the blood stains removed from the Q204 swabbing of the rear gate at the crime scene and from the Q205/6 sock. Traces of EDTA were detected on the stained and unstained cutting from the victim's dress.


In July 1995, Simpson's counsel sent a letter to the FBI requesting all digital data underlying the LC/MS and HPLC testing in the case. The FBI Office of General Counsel (OGC) responded by letter that the underlying digital data had not been saved in the computer. The following day, the OGC clarified this remark in a letter stating that the FBI had saved the digital data underlying its validation studies at the FSRU, but not the data underlying forensic testing by Martz.


On July 25 and 26, 1995, Martz testified in the Simpson case as a witness called by the defense. Martz testified that he came up with the method used to extract blood from the sock and the swatch from the rear gate. He further testified that it took him approximately one week to design the method used for identifying EDTA in blood. He stated that he did his own validation study and did not look at any validation study by the FSRU. He also stated that he was not aware that Quantico had conducted any studies to determine how well his method pulled EDTA out of blood. Later in the testimony, Martz did refer to the work performed by the FSRU.


Martz also answered questions by defense counsel concerning his failure to retain digital data underlying his charts. Martz explained that the raw data is stored on a computer with limited storage space. Martz stated that in this case he printed out the appropriate charts, drew his conclusions, and did not need to look at the data again. Therefore, Martz told us that he permitted the data to be erased. Martz added that there is a tape back-up attachment for this instrument, but it is quite complex. Martz testified that at present there is no way to review this raw digital data; however, it was possible to review the charts representing the digital data or re-run the samples to obtain digital data.


Martz made additional noteworthy statements during his testimony. In particular, when the defense counsel asked whether Martz had decided during a break to become more aggressive in answering questions, Martz responded, I think I decided that I had to be more truthful. I was not telling the whole truth with yes and no answers. . . . I decided that I wanted to tell the whole truth. At another point in the testimony, Martz acknowledged that he had performed analyses using his own blood in May and July 1995, but had not made any notes describing how he conducted the analyses.


After hearing Martz's testimony, Budowle complained to Forensic Science Research and Training Center Chief Kenneth Nimmich that Martz had not properly credited the FSRU for developing the testing procedures. As a result, Nimmich prepared an August 30, 1995, memorandum to Laboratory management that criticized Martz for failing to retain his digital data and for testifying that he did not review or rely upon the FSRU's data and validation study. Nimmich recommended that Martz be orally reprimanded.


In response, Martz wrote two memoranda. In the first, Martz defended his failure to retain digital data by asserting that he had produced a chart of ions of interest whenever coherent data were present. Martz also maintained that he had followed standard CTU practice in allowing the data to be erased as new data was added to the system's limited storage. In the second memorandum, Martz denied that he had misrepresented the involvement of the FSRU. Martz acknowledged that he had worked with members of the FSRU to develop the EDTA test, but added that he had developed and used a negative ion procedure that was different from the procedure developed at the FSRU. Martz also stated that contrary to Nimmich's suggestion, he did not recall receiving a written report of recommended procedures from the FSRU.


After receiving Martz's response, Nimmich wrote a memorandum in which he concluded that Martz had not improperly erased digital data or been less than candid in giving credit to the FSRU. Nimmich noted that Martz had printed out all pertinent data and that it was not practical to save all the digital data. Nimmich also concluded that Martz had given sufficient credit to the FSRU by the completion of his testimony. SAS Chief Randall Murch reviewed the matter and concluded that Martz's testimony raised a performance, not a misconduct, issue. Murch and Deputy Laboratory Director Thompson later counseled Martz about his lack of precision in testifying and his need to give credit to others.


III. Analysis of Whitehurst's Allegations


Whitehurst did not conduct any analyses in the Simpson case and watched only a few minutes of Martz's testimony on television. Whitehurst makes clear that his comments are based on his conversations with others rather than on first-hand knowledge. Based on these conversations, Whitehurst reports the following allegations:


A. The Claim that Martz Committed Perjury by Testifying that He Authored the Testing Procedures


Whitehurst reports that FSRU chemists told him that Martz committed perjury by representing himself as the author of the protocol used to analyze the evidence. According to Whitehurst, these FSRU chemists actually developed the protocol that Martz used.


To evaluate this claim, we reviewed the sequence of events leading to Martz's examination of the evidence. Research chemists at the FSRU in Quantico, Virginia, began preliminary methods development as early as January 18, 1995. Dr. Dean Fetterolf of the FSRU began running samples using liquid chromatograph/mass spectrometry (LC/MS), although, according to Dr. Mark Miller at the FSRU, Fetterolf's initial results were not promising. In the meantime, Dr. Bruce McCord employed ion chromatography (IC) as a detection method. At about the same time, Miller and Martz at the FBI Laboratory in Washington, D.C., also began developing testing methods using the LC/MS method. Throughout this period, Miller and Martz shared information with one another about instrumental parameters, solutions, and preliminary results from testing positive and negative controls (samples with and without added EDTA).


On February 2, 1995, Miller recorded a method for identifying EDTA in dried blood stains using an aqueous extraction followed by Electrospray Liquid Chromatography/Tandem Mass Spectrometry (ES-LC/MS/MS). To test the validity of this method, Budowle prepared 42 dried bloodstain samples for blind testing at the FSRU. On February 9, 1995, using McCord's procedure for preparing the samples, Miller ran the 42 blind stains using ES-LC/MS/MS in the positive ion mode. In his notes of that validation study, Miller reported that this method correctly determined all 42 samples for EDTA. On February 10, 1995, Miller wrote up his method, which entailed aqueous extraction, injections on LC/MS/MS, specific instrumental parameters, and positive ion MS/MS. As part of these efforts, Miller also conducted a recovery study in which he added a known amount of EDTA to a blood sample and then recovered and measured the EDTA from the sample.


Throughout this time period, Miller conferred by telephone with Martz, who continued to test blood samples at the FBI Laboratory. Martz utilized LC/MS/MS in the positive ion mode, but also began using LC/MS/MS in the negative ion mode, a procedure that was not the focus of the work by Miller and McCord. Martz designed this procedure using different instrumental parameters and a different mobile phase than the positive ion mode procedure used at the FSRU. While not as sensitive as positive ion mode procedure, this negative ion mode procedure was more selective for ions of EDTA. On February 8, 1995, Martz conducted his own validation study of this method.


On February 19, 1995, Martz utilized this LC/MS/MS procedure in the negative ion mode to analyze the known samples of preserved blood taken from Nicole Brown Simpson and O.J. Simpson and the bloodstain evidence from the rear gate and socks. Miller assisted Martz in that analysis at the FBI Laboratory. Using this procedure, Martz was able to detect EDTA in the known samples of preserved blood from Nicole Brown Simpson and O.J. Simpson, but not in the bloodstains from the rear gate and socks. These results appear in Martz's report. In the following week, Martz also used the positive ion mode procedure substantially similar to that used at the FSRU to test the known blood samples from Nicole Brown Simpson and O.J. Simpson and to conduct parent-daughter ion experiments with the bloodstain evidence from the gate and socks. Martz also used a High Pressure Liquid Chromatography procedure, which confirmed Martz's reported results.


With this background, we conclude that on balance, Martz's testimony fairly characterized his responsibility for the testing procedures and was not false or perjury. To be sure, early in his testimony, Martz emphasized his own role in developing a set of conditions for using LC/MS/MS in the negative ion mode. He testified that he designed the method over a period of one week and described it as my method. He stated that he conducted his own validation study and did not look at a validation study by the FSRU. He added that if the defense counsel wanted to know about the FSRU's validation study, the defense counsel would have to ask them. He stated that he was not aware of any FSRU study to test the efficiency of his extraction method. Martz also indicated that he was responsible for the method used to extract blood from the sock and gate swatch. At least initially, Martz appeared to ignore the contributions of the FSRU.


Martz corrected that impression later in his testimony, however. Martz testified that the studies that I did and the studies that were done at Quantico demonstrated that one could easily distinguish between preserved and non-preserved blood. He further testified that he ran a blind test on February 8, 1995, and analyzed the results with procedures developed by myself and--at Quantico . . . . Martz later acknowledged that the FSRU had tested the efficiency of his extraction method. Martz further testified that he kept current in the field of mass spectrometry by contact with his peers, noting, [W]e have a staff at Quantico, a lot of Ph.Ds that do a lot of research and we keep in contact with them.


In reaching the conclusion that Martz did not commit perjury, we also considered the comments of the scientists at the FSRU. McCord, Fetterolf, Miller, and Budowle all told us that they did not consider Martz's testimony to be perjurious or inaccurate. Budowle and Miller noted that Martz appeared initially to take credit for work at the FSRU, but later gave credit to the FSRU. Fetterolf and Budowle told us that they viewed Martz's problem as one of presentation and demeanor.


In sum, we find that Martz's testimony was sufficient to communicate his collaboration with the FSRU, as well as his role in developing the particular negative ion procedure used to analyze the evidence. We find that Martz did not testify falsely or commit perjury by claiming that he authored this procedure.


B. The Claim that Martz Misled the Court Concerning the FSRU's Validation Study and Other Matters


According to Whitehurst, scientists at the FSRU also reported that Martz gave misleading testimony concerning the validation studies and other matters.


During his examination by defense counsel, Martz testified that he was aware that the FSRU did something that they called a validation study. The defense counsel asked whether Martz looked at or reviewed these materials. In response, Martz testified that he did not look at those validation materials and noted that the FSRU had not prepared these materials for him.


Our investigation showed that this testimony by Martz was accurate. Miller told us that he and Martz spoke by telephone about the study, but he did not believe that Martz received from the FSRU any written materials about this validation study. Nor, according to Miller, did the FSRU prepare anything else about the validation study for Martz. During our interview, Martz also denied receiving any such materials from the FSRU before testifying. While it seems clear that Martz did not receive or review the FSRU's validation materials, Martz could have been clearer by noting that Miller had advised him of the validation results.


Martz also testified that he conducted his own validation study. Again, our investigation showed that this testimony by Martz was accurate. Martz's notes confirm that on February 8, 1995, he ran samples as a blind test in the negative ion mode. Miller also recalled that Martz conducted his own testing at the FBI Laboratory using negative and positive controls. Miller understood that Martz generated his own samples for these tests.


Whitehurst further reports that Martz was criticized for testifying that the testing method validated itself. The transcript shows that the defense counsel asked who bore responsibility for validating the test, and Martz replied, [t]he test validates itself basically. Standing alone, this statement appears to be nonsensical. However, Martz immediately went on to explain that the process of validation involves determining that the chemical can be extracted, identifying the chemical based on its mass spectrum, establishing the instrumentation, and using standards and controls to confirm the results. Under the circumstances, we do not think that Martz's statement was erroneous or misleading.


Martz also testified that he was not aware that Quantico had conducted a recovery study to determine how well the extraction method pulled EDTA out of blood. Simpson's counsel asked Martz to review that study at the lunch break. After the lunch break, Martz testified that he had telephoned the FSRU and learned that the extraction method removed approximately 93% of the EDTA from the blood sample.


We do not think that Martz misled the trier concerning his knowledge of this recovery study. Miller told us that he had previously informed Martz of the results of the recovery study, but Martz apparently forgot. Miller recalled speaking with Martz at the lunch break and again informing him of those results. Martz likewise told us that he did not recall ever learning about those results until the time of trial. Also consistent with Martz's statement, our review of Martz's notes in the case did not disclose any notes about the recovery study.


Although we conclude that Martz did not mislead the court as suggested by Whitehurst, we think this case illustrates the importance of principal examiners reviewing the work performed by other examiners and researchers. Given the importance of this case and the obvious expertise on the defense side, Martz was surprisingly unprepared for his testimony. For example, Martz was not aware that the FSRU had conducted studies to determine how well his extraction method pulled EDTA out of blood. Nor had he reviewed the charts, graphs, and notes from the validation study conducted by the FSRU. Had Martz conducted this review, he might have side-stepped much of the attack on his method. We think that Martz's failure to prepare explains why he was not clear in describing the validation efforts performed by the FSRU and, by extension, giving credit to the FSRU for that work. Thus, we conclude that one element in Whitehurst's allegations -- that Martz did not review well the work that had been done -- is accurate.


C. The Claims That Martz Misled the Defense Concerning His Erasure of Digital Data and Improperly Erased Digital Data


Whitehurst also reports that Martz was criticized for misleading the Simpson defense. Martz reportedly told the defense that he had erased the digital data underlying the testing, when in fact the FSRU still possessed such digital data.


The evidence does not support Whitehurst's suggestion that Martz misled the defense concerning the digital data. Our investigation showed that on July 9, 1995, Simpson's counsel sent a letter to the OGC requesting the digital data underlying the LCMS and HPLC testing which was done. In response, OGC attorney John Sylvester contacted Martz, who informed Sylvester that he had not saved his digital data. Martz told us that in reporting the absence of this data, he spoke only for himself and did not make any representations on behalf of the FSRU. Accordingly, on July 10, 1995, Sylvester sent a letter to Simpson's counsel stating that the data had not been saved. Sylvester told us that when Budowle received a copy of Sylvester's letter, he notified the OGC that the FSRU had in fact saved its digital data. As a result, on July 11, 1995, counsel in the OGC sent another letter advising Simpson's attorneys that the FBI had saved the digital data for the validation studies but not for the forensic testing. The foregoing shows that Martz truthfully reported that he had not saved the data underlying his own LCMS and HPLC testing. We discern no effort on Martz's part to mislead.


We also find no support for the allegation that Martz improperly erased his digital data. Martz's decision not to retain digital data, while perhaps subject to criticism for tactical reasons at trial, cannot be criticized from a scientific perspective. The American Society of Crime Laboratory Directors/Laboratory Accreditation Board (ASCLD/LAB) provides guidance concerning the documentation requirements for such analyses. Under Essential Criterion of ASCLD/LAB Manual, the Manual provides that case records such as notes, worksheets, photographs, spectra, printouts, charts and other data or records which support conclusions must be generated and kept by the laboratory. ASCLD-LAB Manual at 19 (Jan. 1994) (emphasis added). In its discussion section for that provision, the Manual further provides:


In general, documentation to support conclusions must be such that in the absence of the examiner, another competent examiner or supervisor could evaluate what was done and interpret the data. . . Examples of analytical documentation would include reference to procedures followed, tests conducted, standards and controls used, diagrams, printouts, autoradiographs, photographic, observations, and results of examinations . . . where instrumental analyses are conducted, operating parameters should be recorded.


ASCLD-LAB Manual at 30-31 (Jan. 1994).


Consistent with these guidelines, Martz retained in hard copy form all mass spectra that demonstrated the detection of any significant ions and upon which he based his conclusions. These charts would have enabled a competent examiner to interpret Martz's data and evaluate his conclusions. Martz was not required to retain his digital data. The digital data that Martz allowed to be erased, and which was not otherwise reflected in hard copy, was not material to Martz's conclusions.

In reaching this conclusion, we also recognize that the limitations of electronic storage made it difficult for Martz to retain this data. A mass spectrometer is capable of scanning streams of ions and recording data every one and one-half seconds, resulting in hundreds of scans collected in digital form and relatively few spectra of interest. In view of the amount of data that could be collected, the Finnegan TSQ 700 mass spectrometer used by Martz in the Laboratory had a cumbersome and inefficient long-term storage system. The available magnetic tape back-up system reportedly stored just 90 megabytes per tape and did so sequentially. Given the constraints of the instrumentation, it would have been extremely time consuming to record all of this digital data and difficult to locate and download information when needed.


Although we do not criticize Martz's erasure of the digital data under the circumstances of this case, we are troubled by Martz's other record-keeping practices. Martz testified that he examined his own blood for the presence of EDTA in May and July 1995, but did not make any notes describing how he conducted these analyses. Martz explained at trial that because he examined his own blood in the same way he had examined other samples, he decided not to write down the procedure again. Martz further stated that he did not prepare a report because he considered these runs to be research, not case work. According to Martz, he would not generate a report when he did case-related research if he thought he could readily remember the examination.


Martz's rough notes in this case confirmed the absence of any notes or reports of these examinations. Additionally, Martz failed to number and initial his notes, identify the case number in some notes, or set forth his procedures for some of his testing.


We find the foregoing record-keeping practices to be unacceptable. Martz should have made and retained notes describing his procedures, even if he considered the procedures to be background research and not case work. As a general rule, an examiner should make and retain notes for all work related to any case, but especially work that might be the subject of examination at trial. Further, another examiner should be able to review such notes and have a complete understanding as to all procedures performed in any case. Martz's work in this regard was deficient.


D. Criticism of Martz's Presentation


Whitehurst also reports that scientists in the FSRU were critical of the manner in which Martz testified.


At the outset, we observe that contrary to the suggestion of the defense, Martz's analysis was sound. Martz employed well-established analytical techniques to isolate and identify EDTA in dried blood, and he answered the specific question raised in the case. While Martz came under intense questioning by the defense for not conducting various additional studies, we are not critical of Martz on these grounds. Given an unlimited amount of time and resources, the FBI Laboratory could have conducted all sorts of studies on myriad related and tangential issues. But the reality is that Martz's role was to generate probative information based on the limited samples provided and return the samples for further independent analysis if necessary. He accomplished that task, and it does not appear that any other expert in the case repeated his work and came to any other conclusions.


All things considered, however, the Laboratory would have been better served by assigning another examiner in the CTU to this case. By his own admission during testimony, Martz had many other responsibilities at the time as the Unit Chief and Acting Section Chief. Martz's poor preparation, his lack of a toxicological background, and his maladroitness as a witness were evident when he misstated the value for pi, admitted that he was unfamiliar with the word pharmacokinetics, commented about the need to tell the whole truth, and appeared to boast that he was the foremost expert in EDTA testing. Furthermore, our review of a videotape of portions of Martz's testimony showed that Martz appeared to be unprepared, ill-at-ease, and defensive -- characteristics that undermined his effectiveness as a witness. In a high profile case such as this, consideration should have been given to assigning the case to an examiner with a stronger toxicological background. In this respect, Laboratory management must bear some responsibility for Martz's testimony in this case.


Perceptions by the court or jury as to the credibility of experts are often influenced by factors such as demeanor, presentation, and background. On these dimensions, Martz was not at all impressive.


IV. Conclusion


We find no basis to conclude that Martz committed perjury or misled the trier of facts or defense in the Simpson case. Nor do we conclude that Martz improperly erased digital data underlying his results. We do conclude that because of his lack of preparation, his deficient record-keeping and note-taking practices, and certain aspects of his presentation and demeanor at trial, Martz poorly represented the Laboratory and the FBI in this case.