Department of Health Policy and Management, College of Health Science, and Department of Public Health Sciences, Graduate School, Korea University, Seoul, Korea.
Correspondence to Hae-Young Kim, DDS, PhD. Associate Professor, Department of Health Policy and Management, College of Health Science, and Department of Public Health Sciences, Graduate School, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul, Korea 02841. TEL, +82-2-3290-5667; FAX, +82-2-940-2879; kimhaey@korea.ac.kr
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Asking advice about sample size calculation is one of frequent requests from clinical researchers to statisticians. Sample size calculation is essential to obtain the results as the researcher expects as well as to interpret the statistical results as reasonable one. Usually insignificant results from studies with too small sample size may be subjects to suspicion about false negative, and also significant ones from those with too large sample size may be subjects to suspicion about false positive. In this article, the principles in sample size calculation will be introduced and practically some examples will be displayed using a free software, G*Power.1
Why sample size determination is important?
Sample size determination procedure should be performed prior to an experiment in most clinical studies. To draw the conclusion of an experiment, we usually interpret the p values of significance tests. A p value is directly linked to the related test statistic calculated by using standard errors, which is a function of sample size and standard deviation (SD). Therefore, the results of significance test differ depending on the sample size. For example, in comparison of two sample means, the standard error can be expressed as the standard deviation multipled by root-squared 2/sample sizes (SD / √2n), if equal sample size and equal variance between two groups are assumed. When the sample size is inappropriate, our interpretations based on p values could not be reliable. If we have too small sample size, we are apt to find a small test statistic, a large p value, and statistical insignificance even when the mean difference is substantial. In contrast, larger sample size may lead into a larger test statistic, a smaller p value, and statistical significance even when the mean difference is just trivial. A statistical significance test result may be unreliable when a sample size is too small, and it may be clinically meaningless when too large. Therefore, to make the significance test reliable and clinically meaningful, we need to plan a study with an appropriate sample size. The previous article about effect size in Statistical Notes for Clinical Researchers series provided a more detailed explanation about this issue.2
Information needed for sample size determination
When we compare two independent group means, we need following information for sample size determination.
1. Information related to effect size
Basically we need to suggest expected group means and standard deviations. Those information can be obtained from similar previous studies or pilot studies. If there is no previous study, we have to guess the values reasonably according to our knowledge. Also we can calculate the effect size, Cohen's d, as mean difference divided by SD.
Cohen's d=mean (1)-mean (2)SD
If variances of two groups are different, SD is given as √SD21+SD222 under assumption of equal sample size. To 2 detect smaller effect size as statistically significant, a larger sample size is needed as shown in Table 1.
Table 1
Examples of determined adequate sample size for one-sided tests according to various mean difference, size of standard deviation, level of significance, and power level (allocation ratio N2 / N1 = 1)
Type one error level (α - error level) or level of significance needs to be decided. The significance test may be one-sided or two-sided. For one-sided test we apply Zα for one-sided test, and Zα/2 for two-sided test. Usually for α - error level of 0.05, Zα=0.05 = 1.645 for one sided test and Zα/2=0.025 = 1.96 for two-sided test (Table 2).
Table 2
Significance level for one-sided test, power and corresponding Z values
Power is probability of rejecting null hypothesis when the alternative hypothesis is true. Power is obtained as one minus type two error (1 - β error), which means probability of accepting null hypothesis when the alternative hypothesis is true. The most frequently used power levels are 0.8 or 0.9, corresponding to Z1-β=0.80 = 0.84 and Zβ=0.90 = 1.28 (Table 2).
4. Allocation ratio
Allocation ratio of two groups needs to be determined.
5. Drop out
During the experiment period, some subjects may drop out due to various reasons. We need to increase the initial sample size to get adequate sample size at final observation of the study. If 10% of drop-outs are expected, we need to increase initial sample size by 10%.
Calculation of sample size
When we compare two independent group means, we can use the following simple formula to determine an adequate sample size. Let's assume following conditions: mean difference (mean 1 - mean 2) = 10, SD (σ) = 10, α - error level (two-sided) = 0.05 (corresponding Zα/2 = 1.96), power level = 0.8 (corresponding Zβ = 0.84), and allocation ratio N2 / N1 = 1. The sample size was calculated as 16 subjects per group.
n1=2×(Zα/2+Zβ)2σ2=2×(1.96+0.84)2×102=15.68≈16(mean 1 - mean 2)2
The sample size calculation can be accomplished using various statistical softwares. Table 1 shows determined sample sizes for one-sided tests according to various mean difference, size of standard deviation, level of significance, and power level, using a free software G*Power. The determined sample size of '17' in Table 1 is found on the exactly same condition above. Larger sample size is needed as effect size decreases, level of significance decreases, and power increases.
Sample size determination procedure using G*Power
G*Power is a free software. You can download it at http://www.gpower.hhu.de/. You can determine an appropriate sample size in comparison of two independent sample means by performing the following steps.
Step 1: Selection of statistical test types:
Menu: Tests-Means-Two independent groups
Step 2: Calculation of effect size:
Menu: Determine - mean & SD for 2 groups - calculate and transfer to main window
Step 3: Select one-sided (tails) or two-sided (tails) test
Example 1) Effect size = 2
Two sided (tails) test
Two-sided α - error level=0.05 (one-sided α=0.025)
Power = 0.8
Allocation ratio N2 / N1 = 1.
Appropriate sample size calculated: N1 = 8, N2 = 8.
Example 2) Effect size = 1
One sided (tails) test
α - error level = 0.025
Power = 0.8
Allocation ratio N2 / N1 = 2.
Appropriate sample size calculated: N1 = 13, N2 = 25.
REFERENCES
1. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175-191.ArticlePubMedPDF
2. Kim HY. Statistical notes for clinical researchers: effect size. Restor Dent Endod 2015;40:328-331.ArticlePubMedPMC
Tables & Figures
Table 1
Examples of determined adequate sample size for one-sided tests according to various mean difference, size of standard deviation, level of significance, and power level (allocation ratio N2 / N1 = 1)
1. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175-191.ArticlePubMedPDF
2. Kim HY. Statistical notes for clinical researchers: effect size. Restor Dent Endod 2015;40:328-331.ArticlePubMedPMC
Citations
Citations to this article as recorded by
Denoising Improves Cross‐Scanner and Cross‐Protocol Test–Retest Reproducibility of Diffusion Tensor and Kurtosis Imaging Benjamin Ades‐Aron, Santiago Coelho, Gregory Lemberskiy, Jelle Veraart, Steven H. Baete, Timothy M. Shepherd, Dmitry S. Novikov, Els Fieremans Human Brain Mapping.2025;[Epub] CrossRef
Disinfection of Human and Porcine Corneal Endothelial Cells by Far-UVC Irradiation Ben Sicks, Martin Hessling, Kathrin Stucke-Straub, Sebastian Kupferschmid, Ramin Lotfi Medicina.2025; 61(3): 416. CrossRef
Implementation and Practice of the Antimicrobial Stewardship Program (ASP) in Various Healthcare Settings in Al Ahsa Lorina Ineta Badger-Eme, Promise Madu Emeka, Sayed Abdul Qadar Quadri, Hussein Yousef AlHelal, Abdulrahim Alabdulsal, Mohammed Abdulwahab Buhalim, Omer Abdullah Alrasheed, Muktar Ali AlHelal International Journal of Pharmacology.2025; 21(2): 270. CrossRef
Analysis of the Use of Sample Size and Effect Size Calculations in a Temporomandibular Disorders Randomised Controlled Trial—Short Narrative Review Grzegorz Zieliński, Piotr Gawda Journal of Personalized Medicine.2024; 14(6): 655. CrossRef
Asymptomatic bacteriuria and its associated fetomaternal outcomes among pregnant women delivering at Bugando Medical Centre in Mwanza, Tanzania Colman Mayomba, Dismas Matovelo, Richard Kiritta, Zengo Kashinje, Jeremiah Seni, Seth Agyei Domfeh PLOS ONE.2024; 19(10): e0303772. CrossRef
Obesity Influences T CD4 Lymphocytes Subsets Profiles in Children and Adolescent's Immune Response Rafael Silva Lima, Mayara Belchior-Bezerra, Daniela Silva de Oliveira, Roberta dos Santos Rocha, Nayara I Medeiros, Rafael T Mattos, Isabelle Camile dos Reis, Aiessa Santos Marques, Pedro WS Rosário, Maria Regina Calsolari, Rodrigo Correa-Oliveira, Walder The Journal of Nutrition.2024; 154(10): 3133. CrossRef
Clinical and psychological impact of lip repositioning surgery in the management of excessive gingival display Asmita Dawadi, Manoj Humagain, Simant Lamichhane, Birat Sapkota The Saudi Dental Journal.2024; 36(1): 84. CrossRef
Fifteen-year recall period on zirconia-based single crowns and fixed dental prostheses. A prospective observational study Shahnawaz Khijmatgar, Margherita Tumedei, Guilia Tartaglia, Michele Crescentini, Gaetano Isola, Ernesto Sidoti, Chiarella Sforza, Massimo Del Fabbro, Gianluca Martino Tartaglia BDJ Open.2024;[Epub] CrossRef
A Serious Game for Enhancing Rescue Reasoning Skills in Tactical Combat Casualty Care: Development and Deployment Study Siyue Zhu, Zenan Li, Ying Sun, Linghui Kong, Ming Yin, Qinge Yong, Yuan Gao JMIR Formative Research.2024; 8: e50817. CrossRef
Shifts in Narrative Perspectives Consume Attentional Resources and Facilitate Reading Engagement Jian Jin, Siyun Liu Scientific Studies of Reading.2023; 27(5): 393. CrossRef
Effectiveness of Malaria Free Zone Program on the Knowledge, Attitude and Practice of Malaria Prevention among university students in Conakry, Guinea: Protocol of a Randomized Controlled Trial Aicha Sano, Dhashani A/P Sivaratnam, Norliza Ahmad, Alioune Camara, Poh Ying Lim Malaysian Journal of Medicine and Health Sciences.2023; 19(2): 321. CrossRef
Evaluation of carotid artery Doppler measurements in late-onset fetal growth restriction: a cross-sectional study Gokce Naz Kucukbas, Yasemin Doğan Journal of Surgery and Medicine.2023; 7(10): 673. CrossRef
Assessment of changes in Streptococcus pyogenes levels using N-acetylgalactosamine-6-sulfatase marker and pharyngeal airway space with appliance therapy in mouth breathers – An ELISA-based study Meha Singh, Shivani Mathur, Pulkit Jhingan, Anshi Jain Journal of Indian Society of Pedodontics and Preventive Dentistry.2023; 41(2): 111. CrossRef
Reaffirming Adverse Events Related to Lung Cancer Survivors’ Target Therapies and Their Apparent Effects on Fear of Cancer Progression, Anxiety, and Depression Chu-Chun Yu, Chia-Yu Chu, Yeur-Hur Lai, Chia-Tai Hung, Jui-Chun Chan, Yen-Ju Chen, Hui-Te Hsu, Yun-Hsiang Lee Cancer Nursing.2023; 46(6): 488. CrossRef
Effect of depot medroxyprogesterone acetate on cardiometabolic risk factors among women of reproductive age in Rwanda: A prospective cohort study Evelyne Kantarama, Dieudonne Uwizeye, Annette Uwineza, Claude Mambo Muvunnyi Indian Journal of Medical Sciences.2023; 76: 28. CrossRef
Designing prototype rapid test device at qualitative performance to detect residue of tetracycline in chicken carcass Mochamad Lazuardi, Eka Pramyrtha Hestianah, Tjuk Imam Restiadi Veterinary World.2022; : 1058. CrossRef
To Calibrate or not to Calibrate? A Methodological Dilemma in Experimental Pain Research Waclaw M. Adamczyk, Tibor M. Szikszay, Hadas Nahman-Averbuch, Jacek Skalski, Jakub Nastaj, Philip Gouverneur, Kerstin Luedtke The Journal of Pain.2022; 23(11): 1823. CrossRef
Intravenous Methylprednisolone Pulse Therapy Versus Intravenous Immunoglobulin in the Prevention of Coronary Artery Disease in Children with Kawasaki Disease: A Randomized Controlled Trial Nahid Aslani, Seyed-Reza Raeeskarami, Ehsan Aghaei-Moghadam, Fatemeh Tahghighi, Raheleh Assari, Payman Sadeghi, Vahid Ziaee Cureus.2022;[Epub] CrossRef
Bioactivity effects of extracellular matrix proteins on apical papilla cells Maria Luísa LEITE, Diana Gabriela SOARES, Giovana ANOVAZZI, MON Filipe Koon Wu, Ester Alves Ferreira BORDINI, Josimeri HEBLING, Carlos Alberto DE SOUZA COSTA Journal of Applied Oral Science.2021;[Epub] CrossRef
Mapping QTL associated with partial resistance to Aphanomyces root rot in pea (Pisum sativum L.) using a 13.2 K SNP array and SSR markers Longfei Wu, Rudolph Fredua-Agyeman, Sheau-Fang Hwang, Kan-Fa Chang, Robert L. Conner, Debra L. McLaren, Stephen E. Strelkov Theoretical and Applied Genetics.2021; 134(9): 2965. CrossRef
Oxidized Low-Density Lipoprotein Drives Dysfunction of the Liver Lymphatic System Matthew A. Burchill, Jeffrey M. Finlon, Alyssa R. Goldberg, Austin E. Gillen, Petra A. Dahms, Rachel H. McMahan, Anne Tye, Andrew B. Winter, Julie A. Reisz, Eric Bohrnsen, Johnathon B. Schafer, Angelo D’Alessandro, David J. Orlicky, Michael S. Kriss, Hugo Cellular and Molecular Gastroenterology and Hepatology.2021; 11(2): 573. CrossRef
Comparing 24-hour symptom triggered therapy and fixed schedule treatment for alcohol withdrawal symptoms – A randomized control study Raghuthaman Gopal, Sushith Sugathan Chennatte, Shilpa S. Asian Journal of Psychiatry.2020; 48: 101888. CrossRef
Rejuvenation of three germ layers tissues by exchanging old blood plasma with saline-albumin Melod Mehdipour, Colin Skinner, Nathan Wong, Michael Lieb, Chao Liu, Jessy Etienne, Cameron Kato, Dobri Kiprov, Michael J. Conboy, Irina M. Conboy Aging.2020; 12(10): 8790. CrossRef
Management of haemorrhoids: protocol of an umbrella review of systematic reviews and meta-analyses Min Chen, Tai-Chun Tang, Tao-Hong He, Yong-Jun Du, Di Qin, Hui Zheng BMJ Open.2020; 10(3): e035287. CrossRef
mTOR/HDAC1 Crosstalk Mediated Suppression of ADH1A and ALDH2 Links Alcohol Metabolism to Hepatocellular Carcinoma Onset and Progression in silico Kashif Rafiq Zahid, Shun Yao, Abdur Rehman Raza Khan, Umar Raza, Deming Gou Frontiers in Oncology.2019;[Epub] CrossRef
Enhancing Effect of Elastinlike Polypeptide-based Matrix on the Physical Properties of Mineral Trioxide Aggregate Ji-Hyun Jang, Chung-Ok Lee, Hyun-Jung Kim, Sahng G. Kim, Seung-Wuk Lee, Sun-Young Kim Journal of Endodontics.2018; 44(11): 1702. CrossRef
Statistical notes for clinical researchers: Sample size calculation 2. Comparison of two independent proportions Hae-Young Kim Restorative Dentistry & Endodontics.2016; 41(2): 154. CrossRef
Statistical notes for clinical researchers: Sample size calculation 1. comparison of two independent sample means
Statistical notes for clinical researchers: Sample size calculation 1. comparison of two independent sample means
Examples of determined adequate sample size for one-sided tests according to various mean difference, size of standard deviation, level of significance, and power level (allocation ratio N2 / N1 = 1)
Variation
Group 1 Mean ± SD
Group 2 Mean ± SD
Mean difference
SD
Effect size
Level of significance (one-sided)
Power
Sample size per group
Effect size
40 ± 10
20 ± 10
20
10
2
0.025
0.8
6
40 ± 10
30 ± 10
10
10
1
0.025
0.8
17
40 ± 10
35 ± 10
5
10
0.5
0.025
0.8
64
40 ± 10
39 ± 10
1
10
0.1
0.025
0.8
1571
Standard deviation
40 ± 20
20 ± 20
20
20
1
0.025
0.8
17
40 ± 13.3
40 ± 13.3
20
13.3
1.5
0.025
0.8
9
40 ± 6.7
40 ± 6.7
20
6.7
3
0.025
0.8
4
Level of significance
40 ± 10
30 ± 10
10
10
1
0.05
0.8
14
40 ± 10
30 ± 10
10
10
1
0.01
0.8
22
40 ± 10
30 ± 10
10
10
1
0.001
0.8
34
Power
40 ± 10
30 ± 10
10
10
1
0.025
0.7
14
40 ± 10
30 ± 10
10
10
1
0.025
0.9
23
40 ± 10
30 ± 10
10
10
1
0.025
0.95
27
SD, standard deviation.
Significance level for one-sided test, power and corresponding Z values
Alpha or Beta
Zα or Zβ
0.005
2.58
Significance level = 0.01 for 2 sided test
0.025
1.96
Significance level = 0.05 for 2 sided test
0.05
1.645
Significance level = 0.1 for 2 sided test
0.1
1.28
Power (1 - β) = 0.9
0.2
0.84
Power (1 - β) = 0.8
Table 1 Examples of determined adequate sample size for one-sided tests according to various mean difference, size of standard deviation, level of significance, and power level (allocation ratio N2 / N1 = 1)
SD, standard deviation.
Table 2 Significance level for one-sided test, power and corresponding Z values