Font Size: a A A

The Application Of Multianalysis In The Evaluation Of 《Journal Of Hebei Medicial University》

Posted on:2008-11-07Degree:MasterType:Thesis
Country:ChinaCandidate:Y X ShengFull Text:PDF
GTID:2178360215988982Subject:Epidemiology and Health Statistics
Abstract/Summary:PDF Full Text Request
Objective:《Journal of Hebei Medicial University》is a journal published by Hebei Medicial University. It is an external window of the university in science. Being the organic constitute part of teaching and scientific research of Hebei Medicial University, the basic mission of《Journal》is reflecting fruits of scientific research in this university, accelerating the development of scientific research through this kind of reflection, promoting the growth of teachers, research troops, and our country's medicial science development. Since being started in 1960,《Journal of Hebei Medicial University》's members made great effort to enhance the academic level and editing quality, creating good condition actively and searching for opportunity to make the journal become more and more well-known to attract more authors'contribution. In recent years,《Journal》made great progress in indexes for journal evaluating in《Chinese Science and Technology Journal Citation Reports, CJCR》, and influenced widely in the same kind of local journals. Purposeful analysis and evalution for《Journal》in time is helpful to sum up experiences and recognize its place in the development, making it easy to develop in the future. This article used multianalysis (principal components analysis and factor analysis, weight standard forming calculation and rank sum ratio method) to evaluate《Journal of Hebei Medicial University》in recent 6 years (1999~2004), making self-contr- ast, and trying to search a way for《Journal》's development in the future.Methods:1 Based on the attainable data, 6 years of《Journal of Hebei Medicial University》(1999~2004) were selected as research object.2 All the indexes for journal evaluating were selected from《Chinese Science and Technology Journal Citation Reports》. Based on the attainable data, 13 indexes were selected from《Chinese Science and Technology Journal Citation Reports》, that is: total quoted frequences X 1, inflence factor X 2, annual index X 3, other-quoted ratio X 4, quoted-halflife X 5, original literatures quantity X 6, reference literatures quantity X 7, average-quoted ratio X 8 , average author quantity X 9, area distributing quantity X 10, organization quantity X 11, international paper ratio X 12, fund paper ratio X 13.3 Single index evaluation. Array《Journal of Hebei Medicial University》through total quoted frequences and influencing factors respectively.4 Integrative evaluations4.1 Use principal components analysis and factor analysis to pick up needed principal components and calculate the scores of weighting principal components. 4.2 Use weight standard forming calculation to calculate weights and weighting scores.4.3 Use rank sum ratio method to calculate the weighting rank sum ratio (WRSR) and get the relevant linear correlation equation.4.4 Use weighting average combinative evaluation to get the weights of each method with the scores we got through the three methods above. Calculate the combinative evalution scores, by which to array《Journal of Hebei Medicial University》(1999~2004).Results:1 The results of single index evaluation1.1 Array the 6 years《Journal of Hebei Medicial University》downwards through total quoted frequences. The result is: 2004, 2003, 2001, 2002, 2000, 1999.1.2 Array the 6 years《Journal of Hebei Medicial University》downwards through influencing factor. The result is: 2004, 2002, 2003, 2000, 2001, 1999.1.3 Totally speaking,《Journal of Hebei Medicial University》has developed straightly in recent 6 years, escepcially in 2004. This proved that《Journal of Hebei Medicial University》is becoming more and more approbatory by readers with great effort in recent years.2 The results of integrative evaluations2.1 Principal components analysis and factor analysis2.1.1 Pick up the five anterior common factors, which cumulative contribution rate is 100%. Respectively of contribution rate one by one in order is: 0.253, 0.197, 0.195, 0.181, 0.173.2.1.2 Use Varimax orthogonal rotation. We can regard the first factor as an integrated index indicating external intercommunion level. The indexes loading much on the first factor are: original literatures quantity X 6, reference literatures quantity X 7 and international paper ratio X 12. We regard the second factor as an integrated index indicating all. The indexes loading much on the second factor are: total quoted frequences X 1, inflence factor X 2 . We regard the third factor as an integrated index indicating journal aging and influences. The indexes loading much on the third factor are: quoted-halflife X 5, organization quantity X 11. We regard the fourth factor as an integrated index indicating quoted condition. The index loading much on the fourth factor is: other-quoted ratio X 4. We regard the fifth factor as an integrated index indicating quoting and being quoted condition. The indexes loading much on the fifth factor are: annual index X 3, average-quoted ratio X 8 .2.1.3 Calculate scores of principle components analysis with their weights. Array the samples downwards through weighting scores as follows: 2003, 2004, 2002, 2001, 1999, 2000. 2.2 The results of weight standard forming calculation 2.2.1 Unit the original data and calculate sums respectively. We can get weights for each index as follows:ωX1=0.076,ωX2=0.078,ωX3=0.059,ωX4=0.085,ωX5=0.085,ωX6=0.085, ωX7=0.085,ωX8=0.084,ωX9=0.086,ωX10=0.084,ωX11=0.076,ωX12=0.035,ωX13=0.083。2.2.2 Calculate the weighting indexes values and sums making use of the formula ( X ij = X 'ij×ωj). We can get scores of each sample, and array the samples downwards through it as follows: 2004, 2003, 2002, 1999, 2000, 2001. 2.3 The results of RSR method 2.3.1 Calculate WRSR with the formula as follows: . Array the samples downwards through WRSR as follows: 2003, 2004, 2001, 2002, 1999, 2000.2.3.2 Edit WRSR frequency table. Calculate the cumulative frequences p , which is conversed to probability units(Probit). Set up linear regression equation, taking Probit as independent variables and WRSR as dependent variables. WRSR? =0.190+0.074Probit(F=30.010, P=0.005).2.4 The results of weighting average combinative evaluation2.4.1 Make normality tests to the 3 results we got above, and the results are as follows: principal components analysis: Z =0.395, P =0.998; weight standard forming calculation: Z =0.572, P =0.899; RSR method: Z =0.540, P =0.933. All the results accord normal distribution. Use the probability unit method to standardize all the results.2.4.2 Beforehand test. Doing consistency test to the three groups of standardized data we got the Cronbach'sαcoefficient, which is 0.933, indicating that there is good consistency among the three groups.2.4.3 Contruct the reference sequence for judging. C0 i=(4.605, 3.676, 4.558, 5.017, 6.029, 6.116). The normality test of result for the referenced sequence is as follows: Z =0.499, P =0.965, indicating that the reference sequence according the normal distribution.2.4.4 Calculate the pearson coefficients of the three standardzied data sequences and reference sequence, the results following indicate that the sequences are highly correlated:ρ1 0=0.962,P =0.002;ρ20=0.906, P =0.013;ρ30=0.949, P =0.004.2.4.5 The weight coefficients of each methods in integrative evaluations are:ω1 =0.3415,ω2=0.3216,ω3 =0.3369.2.4.6 Array the samples downwards through the scores of integrative evaluations. The result is: 2004, 2003, 2002, 1999, 2001, 2000.2.4.7 Afterwards tests. Do normality tests to the scores of integrative evaluations. The result showed that it accorded normal distribution( Z = 0.484, P = 0.973). The pearson coefficients among the scores of the integrative evaluation and the 3 groups scores are:ρ1 =0.964, P =0.002;ρ2=0.902, P =0.014;ρ3=0.951, P =0.004; The average correlation coefficient isρ=0.939. These results showed that weighting average combinative evaluation method was correlated highly with the three methods above, so it could indicate the original information and we can get effiective results through it.Conclusion: Array the samples through the scores of combinative evaluation downwards, we can get the result as follows: 2004, 2003, 2002, 1999, 2001 and 2000. This result showed the steady ascending development of recent 6 years of《Journal of Hebei Medicial University》. Particularly in recent 3 years,《Journal of Hebei Medicial University》nearly got a leap of quality.
Keywords/Search Tags:journal evaluation, principal components anlysis and factor analysis, weight standard forming calculation, rank sum ratio method, weighting average combinative evaluation
PDF Full Text Request
Related items