In the process of data analysis, the standards of some analysis methods can be found by just searching. Whether it is word of mouth or default, everyone follows that standard. It’s okay not to look into the source for daily analysis, but in formal academic papers, you need to be responsible for every sentence you write, and every judgment standard should be supported by references, which is one of the meanings of writing references. SPSSAU summarizes some commonly used standard references, come and see if there is anything you need in your paper.
1. Collation of common standard references
For example, everyone knows, "When the VIF values are all less than 10 (strictly 5), it means that the model has no multicollinearity problem and the model is well constructed; on the contrary, if the VIF is greater than 10, the model is poorly constructed."; But when your academic paper When this sentence appears in , it needs to be supported by references in terms of academic rigor. Taking this as a starting point, SPSSAU organizes some references of such common standards for your reference.
Reference list:
1. Normality test standard
"If the sample size is greater than 50, the results of the Kolmogorov-Smirnov test should be used, otherwise the results of the Shapiro-Wilk test should be used."
references:
[1] Chen Pingyan, Huang Zheming. SPSS 13.0 Statistical Software Application Tutorial [J]. People's Health Publishing House, 2005.
[2] Liu Kun. Application of SPSS Statistical Analysis in Medical Research [M]. People's Health Publishing House, 2012.
2、DW值
"If the DW value is around 2 (between 1.7 and 2.3), it means that there is no autocorrelation and the model is well constructed. On the contrary, if the DW value deviates significantly from 2, it means that there is autocorrelation and the model is poorly constructed." References
:
[3] Lu Wendai, Zhu Hongbing. SPSS Statistical Analysis. Fifth Edition [M]. Electronic Industry Press, 2015.
3. Judgment of VIF value
"If all are less than 10 (strictly 5), it means that the model has no multicollinearity problem and the model is well constructed; otherwise, if the VIF is greater than 10, the model is poorly constructed." References
:
[4]J Johnston. Econometric methods 3rd ed.[J]. New York, New York, McGraw-Hill Book Company, 1984, 1984.
[5] Jia Junping, He Xiaoqun, Jin Yongjin. Statistics. 7th Edition [M]. Renmin University of China Press, 2018. [6] Lu Wendai, Zhu Hongbing. SPSS Statistical Analysis. 5th Edition [M] . Electronic Industry Press, 2015.
4. Correlation coefficient of collinearity
"If the correlation coefficient between the analysis items is too high, such as greater than 0.8, it will lead to collinearity"
references:
[7] Xiao Lingyun. Correlation Analysis and Application of Machining Quality Data [J]. Automobile Manufacturing Industry, 2020(6): 3.
[8] Liu Fang, Dong Fenyi. Research on the Diagnosis and Treatment of Multicollinearity in Econometrics [J]. Journal of Zhongyuan Institute of Technology, 2020, 31(1):6.
5. R square related
"A higher R-squared value indicates a better model."
references:
[9] Zhang Hanxia. Scenario Analysis for Linear Regression and Logistic Regression [J]. Automation and Instrumentation, 2022(10):5.
6. Judgment of closeness of relationship
"When |r|≥0.8, it is regarded as highly correlated; when 0.5≤|r|<0.8, it can be regarded as moderately correlated; when 0.3≤|r|<0.5, it can be regarded as lowly correlated; when |r| When <0.3, it means that the degree between the two variables is extremely weak and can be regarded as irrelevant.”
references:
[10] Jia Junping, He Xiaoqun, Jin Yongjin. Statistics. 7th Edition [M]. Renmin University of China Press, 2018.
7. Correlation Coefficient
"When |r|≥0.8 is extremely strong correlation, when 0.6≤|r|<0.8 is strong correlation, when 0.4≤|r|<0.6 is moderate correlation; when 0.2≤|r|<0.4 is weak correlation, When 0≤|r|<0.2, it is a very weak correlation."
references:
[11] Xiao Lingyun. Correlation Analysis and Application of Machining Quality Data [J]. Automobile Manufacturing Industry, 2020(6): 3.
8. T-test effect size
"Criteria for judging effect sizes: Cohen (1988) provides reference standards for judging effect sizes: 0.2 (small), 0.5 (medium), and 0.8 (large)."
references:
[12]Qin Xuefeng, Yang Dongyi, Yang Dongying. The Application and Significance of Effect Size in Independent Sample T Test of Foreign Language Research[J]. 2022(3).
[13] Li Yunming, Yuan Tianfeng, Chen Changsheng, et al. Discussion on Effect Size in ANOVA Model of Medical Research [J]. Journal of Fourth Military Medical University, 2008, 29(3):266-269.
[14] Jacob Cohen.A power primer[J]. Psychological Bulletin, 1992, 112(1):115-159.
[15]Gravetter, Wallnau F , LarryB. Essentials of statistics for the behavioral sciences[M]. Brooks/Cole Publishing Company, 2008.
[16] Jacob Cohen.A power primer[J]. Psychological Bulletin, 1992, 112(1):115-159.
9. Variance effect size index
"The variation caused by the experimental effect accounts for less than 1% of the total as a low level, around 9% as a medium level, and around 25% as a high level."
references:
[17] Quan Chaolu. The meaning and measurement method of effect size [J]. Psychology Exploration, 2003, 23(2):6.
[18]Gravetter, Wallnau F , LarryB. Essentials of statistics for the behavioral sciences[M]. Brooks/Cole Publishing Company, 2008.
[19] Jacob Cohen.A power primer[J]. Psychological Bulletin, 1992, 112(1):115-159.
10. Standards of commonality
"The commonness value of the analysis item is less than 0.3, then the corresponding analysis item should be deleted."
references:
[20] Guo Jinhua, Wu Xing, Tang Yulu, et al. Research on the Measurement Model of the Higher Order Thinking Structure of High School Students—Based on Exploratory and Confirmatory Factor Analysis[J]. Chemistry Teaching, 2019(11):7.
[21] Zhang Wentong. SPSS Statistical Analysis Basic Course (Third Edition College Textbook) [M]. Beijing: Higher Education Press, 2017:106~289.
11. Factor loading value standard
"Item loading values less than 0.4 in factor analysis should be removed."
references:
[22] Guo Jinhua, Wu Xing, Tang Yulu, et al. Research on the measurement model of high school students' chemical higher order thinking structure - based on exploratory and confirmatory factor analysis [J]. Chemistry Teaching, 2019(11):7.
[23] Zhang Wentong. SPSS Statistical Analysis Basic Course (Third Edition College Textbook) [M]. Beijing: Higher Education Press, 2017:106~289.
12. Convergent Validity Criteria
"The AVE value is greater than 0.5, and the CR value is greater than 0.7, indicating good convergent validity."
references:
[24] Guo Jinhua, Wu Xing, Tang Yulu, et al. Research on the Measurement Model of Higher School Students' Chemical Higher-Order Thinking Structure—Based on Exploratory and Confirmatory Factor Analysis[J]. Chemistry Teaching, 2019(11):7.
[25] Fan Jingbo. Differences in Perceived Structure and Level of Residents' Public Service Satisfaction——An Empirical Analysis Based on Exploratory and Confirmatory Factors [J]. Journal of Nanjing University: Philosophy. Humanities. Social Sciences, 2018, (6): 50-52.
13. Fitting index standard
"Chi-square degrees of freedom ratio <0.3, GFI>0.9, NFI>0.9, etc."
references:
[26] Kelly S M , Walton H R . "I'll work out tomorrow": The Procrastination in Exercise Scale[J]. Journal of Health Psychology, 2020(2):135910532091654.
[27] Li Kai. Research on the Performance of Government Purchasing Community Home Care Services [D]. Shanxi University of Finance and Economics, 2022. DOI: 10.27283/d.cnki.gsxcc.2022.000167.
14. Criteria for judging discriminant validity
"The absolute value of the correlation coefficient is less than 0.5 and less than the square root of the corresponding AVE."
references:
[28] Ji Ling, Yao Xiaoyue, Zuo Manfang, Zhou Jia, Wang Xiaomei, Yu Li, Yang Guirong. Reliability and validity test of the Chinese version of chronic disease resources questionnaire in patients with myocardial infarction [J]. China Health Statistics, 2022, 39( 04):588-592.
15. KMO indicator standard
"The KMO value is greater than 0.6, indicating that the correlation between items is high."
references:
[29] Chen Jinyan, Ouyang Jian, Ke Rongrong, et al. Exploratory factor analysis on the behavioral characteristics and influencing factors of blood donors [J]. Chinese Journal of Blood Transfusion, 2018, 31(9):4.
[30] Sun Jia. The Influence of Leadership Emotional Intelligence on Employee Knowledge Sharing [D]. Northeast Normal University, 2022. DOI: 10.27011/d.cnki.gdbsu.2022.001363.
16. Cronbach's reliability coefficient standard
"The alpha coefficient value is preferably greater than 0.8, and 0.7~0.8 is an acceptable range; if the alpha coefficient value is expected to be greater than 0.7, then 0.6~0.7 is also acceptable; if the alpha coefficient value is less than 0.6, you need to reconsider modifying the scale."
references:
[31] Zhou Jun. Questionnaire Data Analysis: Six Types of Analysis Ideas for Cracking SPSS Software. 2nd Edition [M]. Beijing: Electronic Industry Press, 2020: 49~50.
"Cronbach's Alpha coefficient is between 0.7 and 0.9, which indicates high reliability. When Cronbach's Alpha coefficient is lower than 0.35, it is low reliability and needs to be rejected."
references
[32]Xiong Changgen, Liu Jiale. Risk Assessment of Complex Environment Based on Improved Expert Scoring Method[J]. Construction Safety, 2022, 37(11):3.
17. Criteria for determining the number of clusters
"Determining the number of clusters using the number of clusters change point plot."
references:
[33] Wang Bingshen, Wei Yanhua, Zhang Beibei. Improvement of Spectral Clustering Algorithm and Determination of Cluster Number Based on Multivariate Data [J]. Statistics and Decision Making, 2022(012):038.
18. Indicator standards for project analysis
"Item analysis discriminant index CR value, correlation coefficient, homogeneity test, etc."
references:
[34]Wu, Minglong. Pragmatic Questionnaire Statistical Analysis: SPSS Operation and Application[M]. Chongqing University Press, 2010.
19. Coupling coordination degree C value
"Coupling coordination degree C value calculation method."
references:
[35] Cong Xiaonan. The form and nature of the coupling degree model and some misuses in geography [J]. Economic Geography, 2019, 39(04): 18-25. DOI: 10.15957/j.cnki.jjdl.2019.04 .003.
20. Calculation of RI value in AHP
"The calculation method of RI value in AHP."
references:
[36] Hong Zhiguo, Li Yan, Fan Zhihua, Wang Yong. Calculation of Higher Order Average Random Consistency Index (RI) in Analytic Hierarchy Process [J]. Computer Engineering and Application, 2002(12): 45-47+150.
21. Rank and ratio analysis algorithm
"Ranking and Ratio Analysis Algorithmic Procedures."
references:
[37]Xu Yongyong, Sun Zhenqiu. Medical Statistics (Fourth Edition)/Teaching Materials for Colleges and Universities[M]. 2014.
22. Suggested sample size
"The number of effectively recovered questionnaire samples should be at least five times the number of questionnaire items."
references:
[38]Xiong Changgen, Liu Jiale. Risk Assessment of Complex Environment Based on Improved Expert Scoring Method[J]. Construction Safety, 2022, 37(11):3.
2. Types of references
(1) monograph [M], (monograph)
(2) Proceedings [C], (collectedpapers)
(3) Newspaper article [N], (newspaperarticle)
(4) Journal articles [J], (journal)
(5) Dissertation [D], (dissertation)
(6) Technology report [R], (report)
(7) Standard [S], (standardization)
(8) Patent [P], (patent)
(9) Extracted literature in the collection of papers [A], (article)
3. How to get the correct reference format
The format of references contains a lot of content, generally including author name, article title, document type identification, publishing institution, etc. If manual adjustment is complicated and error-prone, you can obtain the correct reference format through CNKI.
(1) Open CNKI, and search for the required documents by searching relevant keywords, document titles, and author names, as shown in the figure below:
(2) Click the "Citation" symbol, and the complete document citation format will appear, just copy it. If you need other format application methods, you can click "Cite More Formats" in the lower right corner to view more citation formats.
