. 2020 Jul 15;10:11647. doi: 10.1038/s41598-020-67321-x

Table 2.

Goodness-of-fit indices of various factor models of the Edinburgh Postnatal Depression Scale.

Lead author, published year	Country	N	Factor model			Goodness-of-fit
Lead author, published year	Country	N	1	2	3	χ²	df	χ²/df	AGFI	RMSEA	CFI	PCFI	TLI	SRMR	kAIC
Previous studies
Cox, 1987	UK	63	1–10			122,849	70	1,755	0.774	0.149	0.766	0.596	0.700	0.083	1,834
Berle, 2003	NOR	411	1–10			122,849	70	1,755	0.774	0.149	0.766	0.596	0.700	0.083	1,834
Gollan, 2017	USA	15,172	1, 2, 6–10			58,964	28	2,106	0.807	0.163	0.813	0.542	0.719	0.075	878
Astbury, 1994	AUS	790	3–5	1, 2, 6–10		83,988	68	1,235	0.845	0.125	0.840	0.635	0.789	0.070	1,795
Matthey, 2008	AUS	238
Phillips, 2009	AUS	167
Swalm, 2010, ante- and post-partum	AUS	4,706	3–5	1, 2, 10		18,227	16	1,139	0.908	0.120	0.932	0.497	0.873	0.083	1,118
Bina and Harrington, 2015	ISR	715	3–5	1, 2, 7–10		74,454	52	1,432	0.841	0.134	0.849	0.613	0.791	0.070	1,540
Cunningham, 2015, on admissions	AUS	571	3–5	1–3, 6–10		89,445	68	1,315	0.831	0.129	0.830	0.627	0.775	0.073	1,801
Cunningham, 2015, on discharged	AUS	543	3–6	6–10	1, 2	24,988	64	390	0.943	0.070	0.953	0.677	0.933	0.039	1,736
Chabrol, 2004	FRN	293	3–7	8–10	1, 2	58,560	65	901	0.873	0.107	0.889	0.642	0.846	0.081	1,770
Small, 2007	AUS	1,168	3–7	9, 10	1, 2, 8	103,127	64	1,611	0.797	0.143	0.804	0.572	0.724	0.082	1,815
Bowen, 2008^a	CAN	402	3–5	10	1, 2, 8	51,766	22	2,353	0.784	0.172	0.851	0.446	0.716	0.095	1,348
Tuohy and McVey, 2008	UK	440	3–5	7–10	1, 2	20,300	48	423	0.945	0.073	0.959	0.639	0.938	0.037	1,486
Kwan, 2015	SGP	920
King, 2012	USA	169
Reichenheim, 2011	BRA	811	3–5	7–10	1, 2, 6	38,341	64	599	0.915	0.087	0.927	0.659	0.898	0.069	1,750
Kubota, 2014	JPN	690	3–5	7–9	1, 2	17,987	34	529	0.937	0.082	0.958	0.582	0.932	0.038	1,348
Takehara, 2018	JPN	1,311	3–5	7–9	1, 2	17,987	34	529	0.937	0.082	0.958	0.582	0.932	0.038	1,348
Chiu, 2017	USA	515	3–6	7–9	1, 2	22,573	48	470	0.939	0.077	0.952	0.634	0.927	0.041	1,598
Coates, 2017	UK	12,166	3–6	7–10	1, 2	24,917	64	389	0.945	0.070	0.953	0.677	0.934	0.040	1,736
Pop, 1992	NED	293	3–6	7–10	1, 2	24,917	64	389	0.945	0.070	0.953	0.677	0.934	0.040	1,736
Kozinsky, 2017, antepartum	HUN	2,967	4, 5	8, 9	1, 2	3,592	12	299	0.973	0.061	0.988	0.395	0.971	0.022	909
Kozinsky, 2017, postpartum	HUN	714	4, 5	3, 6, 10	1, 2	10,528	22	479	0.951	0.078	0.965	0.505	0.933	0.039	1,357
Flom, 2018	MEX	628	4–6	3, 7–9	1, 2	43,164	48	899	0.885	0.106	0.907	0.605	0.861	0.059	1,619
Current study
3-factor	JPN	91,063	3–6	7, 9, 10	1, 2	11,145	48	232	0.972	0.054	0.974	0.649	0.961	0.033	1,546
2-factor	JPN	91,063	3–10	1, 2		64,977	68	956	0.853	0.110	0.876	0.662	0.837	0.061	1,776
1-factor	JPN	91,063	1–10			122,849	70	1,755	0.774	0.149	0.766	0.596	0.700	0.083	1,834
Recommended						Lower		Lower	≥ 0.95	≤ 0.06	≥ 0.95	Higher	≥ 0.95	≤ 0.08	Lower

df degree of freedom, AGFI adjusted goodness-of-fit index, RMSEA root mean square error of estimation, CFI comparative fit index, PCFI parsimonious CFI, TLI Tucker–Lewis index, SRMR standardised root mean square residual, kAIC Akaike information criterion/1,000.

^aNegative error variance appeared in item 10.