Tuesday, June 08, 2010

Evidence and Admissibility Part 1

Evidence and Admissibility Part 1

The analysis that can never stop relates to issues associated with evidence and admissibility. We need thought provoking discussions to inform our knowledge. Also, we need to exercise our understanding of the work we are doing and whether it meets an acceptable criteria in the practices and procedures we follow. Moroever we should challenge perceptions we hold to discover where things could go wrong and make revelation of them to see how best they can be put right.

The discussion below contains references to Statutory Law and Case Law. I am not legally qualified and do not hold myself out to be giving legal advice. The legal references that have been used are there to help identify requirements that are in place or to merely demonstrate an interpretation.

STATEMENTS, TOOLS & PROCEDURES
The independence, objectivity and impartiality of a witness or expert can only be seen if each are responsible for producing their own research, analysis and reports and their own exhibits and the content in them unfettered from control by others who are not giving the evidence. Problemical is that any desire to achieve these laudable principles can be thwarted due to deficiencies in the system.

Using mobile phone examination tools to acquire data, that data maybe subjected to the limited capability of the tool that extracted the data in the first place from the target device (the "exhibit"). The accuracy of the harvested data can again have limitations if the tool that extracts data is equally set to work to translate the data. Most of 'the' tools do this. When I put the word 'the' in parenthesis I do so because I am of course referring to mobile phone tools and not every tool. Notwithstanding the aforementioned limitations, a further limitation can occur from which a tool outputs the data into datafields. This largely leaves the examiner to believe, any data that does not appear or cannot fit into a datafield, it is not (worthy) evidence for inclusion and doesn't get served. Extracted data that are not the subject of translations remains in a raw state in the electronic file and doesn't get served. If the examiner doesn't know what data to look for or does not understand the relevance of the data this too can contribute to evidence not seeing the light of day. If this is not revealed in the report/statement, and left unchallenged, it corrodes the principles and right to fair trail by diluting revelation in the criminal justice system.

It can also be the case that where the report/exhibit design and datafields for cell site analysis evidence are predefined and examiners/experts are told the format they must follow, again this leads to corruption of the principles mentioned at the outset. If the format is predefined it causes the examiner/expert to look only at a certain set of criteria and only the data for that criteria; other data that should have been considered gets ignored, and again does not see the light of day. Furthermore, it prevents the examiner/experts from corroborating facts prior to disclosure.

What observations might be drawn from the views above? One observation might be to look at the Industry Standards, Technical Standards (GSM/3GPP, ETSI/ITU/ISO, proprietory, etc) and QA Standards (ISO etc) to determine discovery and revelations that should be made. For instance, using the example of the proposed QA Standards of ISO17025 and ISO17020 the principles in them are drawn from ISO9000, an international QA Standard having principles in it that originated from the British QA Standard BS5750. So there is as an historical reference and appreciation to be aware that Britain is fully aware of using a quality approach when dealing with 'technical' material and the importance of its revelation. However, the different between ISO17025/ISO17020 and ISO9000/BS5750 is that the former are said to be directly applicable for forensic evidence where as the latter have never been directly applicable for forensic evidence, but have often been suggested as achieving a criteria sufficient for evidence.

More importantly, to this discussion, would the tools used for examination fail the crietria set out in ISO17025 and ISO17020? An example to look at might be the QA criteria relating to "non-conforming" product. If the examination tool does not possess the capability to know if the product (the "exhibit") is non-conforming or by use of that tool it introduces non-conformance into the product (the "exhibit"), then the tools output may fail the test. The analysis of the failure thus requires to be determined: to define whether the coding in the tools have been written incorrectly or the coding incorporates commands of a generic nature - meaning device-specific commands may not have been considered or used? The latter might cause data changes residing in the handset. Changes to data and failure to reveal that prior to disclosure undermines the discovery to be understood about served data and can affect a tribunal's understanding of the "defendant's behaviour in relation to the data" (s129 CJA (Criminal Justice Act) 2003). Under s69 of the Police and Criminal Evidence (PACE) Act 1984 (repealed) that occurrence would have been less likely because s69 compulsorily required prior consideration and affirmation before admissibility and allowed transparency upto the point that enabled probative enquiries and tests of the system (in some cases) to seek validation about a system's performance.

Evidence submitted in the past (pre s69 PACE repeal) came with a meaningful statement that was naturally attached with the evidence; post s69 PACE there is no equivalent technical statement of any compliance whatsoever, unless it is voluntarily made but simply not referenced to s69. When such a statement in the past was made it contained references to affirming the position required by s69 PACE 1984 as to the operation of the computer/s and affirmation (certificate) required by other provisions.

One might be forgiven for thinking the Police when they present evidence are excluded from making at least some declarations about the operational performance of their computer/s and program/s generating evidence. They are not excluded from doing so, there is simply no enforcement to include good QA declarations in their statements/reports about evidence they generate - whether obtained from mobile phones, computers, CCTV/Video and so on. In contrast, currently mobile operators producing billing records etc (served as evidence generated from a computer and program/application) still make a QA reliability statement that the computer that generated the evidence operated properly and if not it would not as such cause alteration to the state of the business record (data). So why shouldn't the Police and examiners make the same statements? Afterall they do use computers and programs to extract and harvest data from an exhibit and it is that data from the computers that the court sees.

CONTROLING REVELATION
The discussion, having dwelt on the subject of identifying some aspects of limitations that can occur with examiner skillsets, flawed tools, contaminated evidence, conflicts with standards and lack of appropriate procedural requirements, there are still the issues of 'control' of revelation as to how evidence is presented and what might influence that to happen?

Where examiners are forced to extract, harvested, examine and report evidence with preset requirements for revelation that may not fully corroborate "behaviour" when compared with the data that has been revealed under any enforced procedure, it may amount to undue preference or undue discrimination. The words 'preference' and 'discrimination' do not attract prohibition. Instead, the emphasis is on 'undue', which in dictionary terms can mean "improper", "unreasonable" and, in particular, "excessive". That emphasis might be applicable where commercial contracts, such as public sector contracts for the employment of outsourced services occurs that may have legal requirements, either express or implied, associated with them, that require the contracted party to use discrimination when revealing data by predicting only particular data to be available and be served (disclosed).

The matters of undue preference and undue discrimination have been dealt with by the House of Lords previously. Lord Keith and Lord Merriman outlined definitions in two separate actions involving the same case - South of Scotland Electricity Board -v- The British Oxygen Company - that concerned disputes with the provision of costs and supply of electricity as to how the meaning of 'undue' may have influence. These cases may provide useful pointers where controlled evidence is being used. Lord Keith in the first action ( [1956] 1 W.L.R. 1069 ) stated that the word "undue" encompassed not only illegitimate reasons but could also mean "excessive". In the later action ( [1959] 2 All E.R. 225 ) Lord Merriman addressed the matter in three parts. The first is use of the phrase "shall not exercise any undue discrimination" had to be considered as a whole. Secondly, making reference to objectives he identified "a fair distribution of the cost of supplying electricity as between one class of users and another is not to be left out of consideration any more...." followed by the use of the incisive comments for the third part "....'than any other circumstance which would affect mens' minds."

If the above principles are applied to evidence, then it would suggest the need to re-instate encumbent obligations (like s69 PACE) requiring performance of fair acquisition of data and fair distribution of evidence for delivery to a class of users (eg the Courts, Prosecution and Defence) and to protect against blind-siding and stumbling blocks that may cause undue discrimination. In addition, public sector contracts may well benefit from containing a contractual clause protection mechanism to enable the immediate use of the law to act as a 'shield and not a sword' when a penalty is threatened or issued against the outsourcer where the outsourcer has identified that the data under the schedules supporting the contract does not permit other important data to be incorporated into the datafields of the template that is being enforced to be used, thus prevents revelation. Such omission of data prevents the Prosecution from meeting their obligations under the Golden Rule enunciated by the noble Lord, Lord Bingham in the case of R .v. C & H (February 2004) to conduct a fair risk assessment when in possession of all the evidence and the facts in order to determine the risk, if any, that the evidence poses. The impeccable logic behind the requirement: if there is fundamental misunderstanding about the evidence and there are too many twist and turns being used to skew the path to appropriate discovery and revelation, that may interfere with others (who are not technical) in allowing them to meet their duty required by the statutory provisions, e.g. s9(2) Criminal Procedures and Investigations Act 1996?

No comments: