EVALUATION OF ARGUMENTS IN RESEARCH REPORTS

Some arrthors on research methodology are of the opinion that research reports are based on the logic of reasoning and that such reports comm~rnicate with the reader by advancing logical, coher-ent arsrments (Bvhme, 1975:206; Mouton, 1996.69). This view implies that researchers dm?v certain conclrrsions and that such conclr~sions be jrrvtlfied by means of reasoning (Doppelt, 1988:105; Giere, 1984:26; Harre, 1965:ll; Lehrer & Wagner: 1981:3; Pitt, 1988:7). Thestnrctzire ofresearch reports, therefore, consists mainly ofconclrrsions andthe reasons forsrrch conclzrsions (Booth, Colomb & CENiams, 1995:97). From this, ithecomes clear thatjrrstification by means of reasoningconstitrrtes a standardprocedrrre in research and research reports.

-What are thestandards that thereasoningin research reports in the htrman sciences .shorrldrneet.~ .Ho+vcan research reports in thekrrmansciences be eval~rntedaccordingto these standards?
In answering thefirst question, the logical demand.^that are made on reasoning in research are investigated.From these demands the acceptability of the statements, relevance andsrrpport oft he premise.^to the conclrrsion are set as standards for reasoning in research.
In answering the second question, a research article is rrsed to demonstmte how the macro-and micro-arguments of research report.7 can be evalrrateriaccorcling to thesestandar1:l.s.

I N T R O D U C T I O N A N D P R O B L E M STATEMENT
Some authors on research methodology are of the opinion that research reports are based on the logic of reasoning and that such reports communicate with the reader by advancing logical, coherent arguments (Bohme, 1975:206;Mouton, 1996:69).This view implies that researchers draw certain conclusions and that such conclusions be justified by means of reasoning (Doppelt, 1988: 105;Giere, 1984:26;Harre, 1965:ll;Lehrer & Wagner, 1981 :3;Pitt, 1988:7).The structure of research reports, therefore, consists mainly of conclusions and the reasons for such conclusions (Booth, Colomb &  Williams, 1995:97).From this, it becomes clear that justification by means of reasoning constitutes a standard procedure in research and research reports.Various theoretical perspectives on research support this view.Rossouw (1993:96) describes scientific findings as "findings that, in principle, can stake a claim for the greatest possible measure of intersubjective consensus" Research should be executed in such a way that its findings would be on a par with and stand the critical testing of other experts in the same domain.This, in turn, implies that research reports should be aimed at a specific research community and that the researcher should advance arguments to convince this research community of the validity of the conclusion of theresearch.
most textbooks on research methodology (Bums &  Grove, 1993:647; Creswell, 1994:193; LoBiondo-Wood & Haber, 1994:441/48 I), does not focus on the arguments used in research.This means that the criteria for the evaluation of research reports are comprehensive and that these criteria must be very specific for each type of research (for example, qualitative or quantitative research).Should the evaluation of research reports focus on arguments and logic, one universal set of standards against which to evaluate all types of human-science research reports couldpossibly be accepted.
As arguments form the basic structure of research reports and are probably also important in the evali~ation of research reports in the human sciences, the following questions are relevant to this paper, namely: What arc the standards that the reasoning in research reports in the human sciences should comply with?How can research reports in the human sciences be evaluated according to these standards'?
In answering the first question, the logical demands that are made on reasoning in research are investigated.From these demands the standards for reasoning in research are formulated.

Despite the fact that the logic of research is based on
In answering the second qiiestion, a research article is reasoning, that the justification of research findings by used to demonstrate how the macro-and micromeans of reasoning appears to be a standard procedure arguments of research reports can be evaluated and that the structure of research reports consists of according to these standards.arguments, the evaluation of research, as described in BACKGROUND following as requirements for solid reasoning: I11 order to describe the standards for the evaluation of research reports and the explanation of how these standards could be applied in evaluating research reports, it would be necessary first to define the concepts macro-argument and micro-argument.
The components of the macro-argume~~t constitute the research decisions taken, with the research problem being the first conlponent of the research report.The principal aim of research is to find a solution to the research problem.For this reason, the researchpurpose and objectives take their place as the second component of the research report.The research problem determines the nature of evidence required (Mouton, 1996:72;Van Veuren, 1996:13).
Evidence is, therefore, gathered that could best solve the research problen~.Evidence is furnished in the fonn of empirical and theoretical statements arising from the conceptualisation and operationalisation of research.As a result, conceptualisation and operationalisation are also listed as components of research.Conceptualisation refers to the defining of the key concepts of the research, as well as to the integration of the research findings into the greater conceptual framework or body of accepted knowledge.Operationalisation, on the other hand, refers to linking the key concepts to the actual phenomenon to be studied.In addition; it inlplics methods of data collection, data analysis, data interpretation and the choice of the population and sampling as well as measurenlents for validity and reliability.The conclusions arrived at by the researcher are based on the evidence collected, with the result that the conclusions constitute the last coillponent of research.
Except for the so-called macro-argument that refers to the logical coherence between the components of research, so-called "micro-arguments" are also advanced within each component in support of the macro-argument.The motivation of the research problem really is a micro-argument within the component "problems statement" of the macroargument.The research report, therefore, comprises inultiple arguments of what are typified as the macroand micro-arguments of research in the present paper.

STANDARDS FOR THE EVALUATION OF ARGUMENTS IN RESEARCH REPORTS
In order to be able to evaluate these macro-and microarguments in research reports, it is necessary to set standards to which the reasoning in research must comply.The logic, to which research reports must comply, is the logic of reasoning.For this reason, it is necessaly to investigate the demands of logic made on reasoning in research in order to lay down standards for the evaluation of research reports.Govier (1991:68) and Van Veuren (1991:34) elect the The acceptability ofthe statements.
-The relevance of the statements to the conclusion.Adequate support forthe conclusion.
The above requirements for solid reasoning were used as a springboard for investigating and laying down the standards for reasoning in researchreports.Following, each of the said three requisements for solid reasoning were investigated from a perspective of logic and contextualised within the research process.From this investigation the following three standards for the evaluation of argument in research reports were described, namely:

The acceptability of the premises and statements
It becomes clear from the investigation that the first standard the statements in a research report must comply with is that of acceptability.The acceptability of statements means that the reader could, on rational grounds, deem such statements to be true.For this reason, there must be a reason why the statements could be deemed acceptable.The reasons that have weight in research involve the researchers ability to confirm that a statement has been observed empirically and that it is in line with existing theoretical frameworks.The researchers, therefore, make especially empirical and theoretical statements in a bid to convince their readers ofthe acceptability oftheir statements.

The relevance of the statements to the conclusion
The second standard to be met by arguments in research can be formulated as follows: The premises in arguments in research ought to have relevance to the conclusion.Micro-arguments especially those based on elnpirical data and existing theoretical frameworks, ought to show the relevance of the premises to the conclusion.

Adequate support for the conclusion
A third standard to be met by the arguments in research reports is that the premises must lend adequate support to the conclusion and that the conclusion innst, in turn, be qualified.

THE EVALUATION OF A RESEARCH ARTICLE ACCORDING TO THESE STANDARDS
To answer the second question of the paper the following research article will be used to demonstrate how the standards can be applied to evaluate the arguments in research reports.

Research problem and research questions
The problem statement is reflected in the first and third paragraph under the heading "background" and reads as follows:

Purpose and objectives
The objectives of the sntdy are described m the second paragraph under the head~ng background and read as
Only the third objective is relevant atid supportive to t l ~e problem statement.The first and second obiectives are ilready implied by the third objective.1t wili be part of a research design to determine the effect of developmental instruction strategies on the cognitive development scores of freshman nursing students and should therefore be part ofthe designdescription.
The fourth objective is tiat relevant nor, does it suppoit the problem statement.It can therefore be omitted from this research article.If the authors want to remain with this objective, the problem statement should be adapted by adding arguments with regards to grade point average and abilities.
Froill the before ilientioned evaluation one should recommend a change in the objectives to read as follows, in order to be logical-coherent to the current problem statement: The purpose/objective of the study is to determine the effect of developinental instiuction strategies on the cognitive development scores of freshman nursing students.

Conceptualisation
The conceptuaiisatio~i of the article is described by the heading theoretical fiainework and literature review and canbc sunmiarised as follows: The perspectives of ' Peny (1970)  HEALTH SA GESONDHEID Vol. 5 NO. 1 -2003 Results of previous research based on Peri-yk perspective,^ ai*e described.

Operationalisation
The operationalisation is discussed under the headings "methodology and data analysis" and consists of research design, population and sampling, datacollection, data analysis as well as validity and reliability.
A quasi-experimental design is relevant and will support the suggested revised problem statenlent and objective ofthe study.
The descriptive design is however not relevant to any conlponent of the study.No descriptive data was given as part ofthe results ofthis research.
Froin the before mentioned evaluation it could be suggested to delete the descriptive design in this study.

Population and sampling
A convenience sample of intact classes, scheduled by the university registrar, was used.One section received the experimental treatment (developmental instruction strategies) and the four other groups of equal size received traditional teaching approaches.Tlze developnzent instruction ~v a s organised arouizd tlze coizcepfs of challeiige andsuppoist jSarifo~-d, 1967).It was flze task of tlze teacher to provide tlze appropriate balaizce ofcliallerzge orsupport fop!-onzotegro~vtlz.
Apart from the logical problem already mentioned, data collection by the LCQ is relevant and supported by the previous components ofthe research article.
The one minute paper adapted by Gross (1981), however, is a not relevant to the research problem or objective of this research and should be omitted.It will be relevant to a research objective to investigate students experience of developmental instruction strategies.

Findings
Tlze arlalysis of flie Septerizbel; May arid Deceiliber scorAes denlonstrated lzo sigriifcalzt difSe~,elzces behveeir the groups.The niearz cogr~itive developnie~zt for all fieslznzelz nul-sing studerzts of each of tlze three testi~igs was 3.4, 3.66 and 3.61 In order to be relevant and to support the previous components the co~~elation data should be omitted from this research article.