中华核医学与分子影像杂志
中華覈醫學與分子影像雜誌
중화핵의학여분자영상잡지
Chinese Journal of Nuclear Medicine and Molecular Imaging
2014年
5期
401-404
,共4页
统计学%期刊%核医学%中国
統計學%期刊%覈醫學%中國
통계학%기간%핵의학%중국
Statistics%Periodicals%Nuclear medicine%China
目的 了解《中华核医学与分子影像杂志》(CJNMMI)中论著的统计学方法应用状况,为进一步提高医学期刊的质量提供依据.方法 利用多阶段随机抽样方法,根据杂志每年的期刊数及论著数抽取2007至2011年发表在CJNMMI上的论著,分别从统计设计、统计描述、统计推断和统计表达4个方面分析其统计学方法的应用及误用情况.结果 共抽取论著116篇.统计设计方面,99.1%(115/116)的论著设立了对照且应用合理,使用随机和盲法的论著分别为16.4%(19/116)和8.6%(10/116),其中有6篇出现随机误用,3篇出现盲法误用;所有论著都应用统计描述方法,使用较多的指标为相对数、均数和标准差,正确应用率均高于90.0%,45.7%(53/116)的论著使用了统计表且无误用;71.6%(83/116)的论著应用了统计推断方法,其中20.5% (17/83)为高级统计推断方法并均应用合理,统计推断误用率为29.7%(27/91),其中方差分析、t检验和x2检验等简单统计推断方法误用较多,分别达到27.8% (5/18)、25.5% (12/47)和24.0%(6/25);统计表达方面,98.0%以上的论著给出了样本量和统计量,有60.3%(70/116)出现误用,主要为未给出样本量确定依据(55.2%).结论 CJNMMI论著中统计学方法应用情况较好,其中以对照、统计描述、高级统计推断方法的应用质量较高,但需加强随机、盲法、t检验、x2检验、方差分析和统计表达等的应用质量.
目的 瞭解《中華覈醫學與分子影像雜誌》(CJNMMI)中論著的統計學方法應用狀況,為進一步提高醫學期刊的質量提供依據.方法 利用多階段隨機抽樣方法,根據雜誌每年的期刊數及論著數抽取2007至2011年髮錶在CJNMMI上的論著,分彆從統計設計、統計描述、統計推斷和統計錶達4箇方麵分析其統計學方法的應用及誤用情況.結果 共抽取論著116篇.統計設計方麵,99.1%(115/116)的論著設立瞭對照且應用閤理,使用隨機和盲法的論著分彆為16.4%(19/116)和8.6%(10/116),其中有6篇齣現隨機誤用,3篇齣現盲法誤用;所有論著都應用統計描述方法,使用較多的指標為相對數、均數和標準差,正確應用率均高于90.0%,45.7%(53/116)的論著使用瞭統計錶且無誤用;71.6%(83/116)的論著應用瞭統計推斷方法,其中20.5% (17/83)為高級統計推斷方法併均應用閤理,統計推斷誤用率為29.7%(27/91),其中方差分析、t檢驗和x2檢驗等簡單統計推斷方法誤用較多,分彆達到27.8% (5/18)、25.5% (12/47)和24.0%(6/25);統計錶達方麵,98.0%以上的論著給齣瞭樣本量和統計量,有60.3%(70/116)齣現誤用,主要為未給齣樣本量確定依據(55.2%).結論 CJNMMI論著中統計學方法應用情況較好,其中以對照、統計描述、高級統計推斷方法的應用質量較高,但需加彊隨機、盲法、t檢驗、x2檢驗、方差分析和統計錶達等的應用質量.
목적 료해《중화핵의학여분자영상잡지》(CJNMMI)중론저적통계학방법응용상황,위진일보제고의학기간적질량제공의거.방법 이용다계단수궤추양방법,근거잡지매년적기간수급론저수추취2007지2011년발표재CJNMMI상적론저,분별종통계설계、통계묘술、통계추단화통계표체4개방면분석기통계학방법적응용급오용정황.결과 공추취론저116편.통계설계방면,99.1%(115/116)적론저설립료대조차응용합리,사용수궤화맹법적론저분별위16.4%(19/116)화8.6%(10/116),기중유6편출현수궤오용,3편출현맹법오용;소유론저도응용통계묘술방법,사용교다적지표위상대수、균수화표준차,정학응용솔균고우90.0%,45.7%(53/116)적론저사용료통계표차무오용;71.6%(83/116)적론저응용료통계추단방법,기중20.5% (17/83)위고급통계추단방법병균응용합리,통계추단오용솔위29.7%(27/91),기중방차분석、t검험화x2검험등간단통계추단방법오용교다,분별체도27.8% (5/18)、25.5% (12/47)화24.0%(6/25);통계표체방면,98.0%이상적론저급출료양본량화통계량,유60.3%(70/116)출현오용,주요위미급출양본량학정의거(55.2%).결론 CJNMMI론저중통계학방법응용정황교호,기중이대조、통계묘술、고급통계추단방법적응용질량교고,단수가강수궤、맹법、t검험、x2검험、방차분석화통계표체등적응용질량.
Objective To assess the correctness of statistical methods applied in the research papers of Chinese Journal of Nuclear Medicine and Molecular Imaging (CJNMMI) and to provide suggestions for further improving.Methods Papers were obtained from CJNMMI by multi-stage sampling method between 2007 and 2011 according to the number of journals and articles each year.The statistical methods were reviewed in the aspects of statistical design,description,inference and expression.Results A total of 116 papers were collected.For the statistical design,99.1% (115/116) used the control group and all were applied properly.The methods of randomization and blinding were used in 16.4% (19/116) and 8.6% (10/116) of papers,respectively,with 6 and 3 papers applied incorrectly.Statistical description was used in all papers with an accurate rate over 90.0%.The index for statistical description mainly included relative number,mean and standard deviation.The frequency for the usage of statistical tables was 45.7% (53/116) and all were applied correctly.In 71.6% (83/116) of papers,statistical inference methods were used,including t test,x2 test and analysis of variance.The overall incorrect rate was 29.7% (27/91),with 27.8% (5/18),25.5% (12/47) and 24.0% (6/25) for analysis of variance,t test and x2 test,respectively.The frequency for the usage of advanced statistical inference methods was 20.5% (17/83) and all were applied correctly.For the statistical expression,98.0% of papers provided the sample size and statistics.The incorrect rate of statistical expression was 60.3% (70/116) and majority of the mistakes was the lack of basis for determining the sample size.Conclusions The overall correctness of statistical methods applied in the papers of CJNMMI was good,especially in statistical design,statistical description and advanced statistical inference methods.However,quality application skills need to be improved in areas such as methods of randomization,blinding,basic statistical inference and statistical expression.