Perceived Active Control over Online Health Information: Underlying Mechanisms of Mode Tailoring Effects on Website Attitude and Information Recall

Figures & data

Figure 1. Mode-tailored website with text and video selected (top) and non-tailored standardized website with text, visuals and video (bottom).

Table 1. Descriptive statistics and correlation matrix of study variables (N = 392)

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age	-
2. Gender ^a	.17**	-
3. Education	−.34***	−.07	-
4. Cancer self ^b	.29***	.04	−.07	-
5. Cancer other ^b	.02	−.09	.07	.01	-
6. Internet use	−.09	.08	.04	−.08	.02	-
7. Internet efficacy	−.22***	−.09	.13*	−.05	−.05	.12*	-
8. Tailored condition	.01	.02	.01	−.00	−.04	−.03	.06	-
9. Active control	.01	−.04	−.02	−.00	−.01	.02	.21***	.11*	-
10. Cognitive load	.14**	.20***	−.12*	.01	−.12*	−.00	−.23***	.01	−.45***	-
11. Personal relevance	.15**	−.07	−.12*	.09	.01	−.06	.11*	.06	.61***	−.31***	-
12. Website engagement	.13**	.02	−.09	.06	.12*	.00	−.17**	−.06	.15**	−.02	.11*	-
13. Website attitude	.04	−.06	−.08	.01	−.00	.00	.24***	.07	.63***	−.30***	.57***	.08	-
14. Information recall	−.07	−.19***	.16**	.00	.14**	.00	.04	−.10	.29***	−.34***	.20***	.28***	.23***	-
Mean	54.71	.53	4.32	.11	.87	19.83	5.58	.47	5.55	2.28	4.68	361.27	5.13	25.82
SD	14.50	.50	1.76	.31	.33	13.63	1.14	.50	1.19	1.33	1.43	424.82	1.06	22.07

Note. ^aGender was dummy coded into 1 = male, 0 = female. ^bExperience with cancer was dummy coded into 1 = has (had) cancer/knows someone with cancer, 0 = has (had) no cancer/does not know someone with cancer.

* p <.05. ** p <.01. *** p <.001.

Table 2. Website use patterns of the mode-tailored and standardized website (N = 392)

		Mode-tailored website (n = 183)		Standardized website (n = 209)		Total (n = 392)
		M	SD	M	SD	M	SD
Time spent on website		334.15	356.08	385.02	476.54	361.27	424.82
Number of web pages visited		2.85	0.36	2.88	0.33	2.87	0.34
Number of (re)visits webpages		11.31^a***	5.78	6.43	4.36	8.71	5.62
	Number of visits Homepage	4.93	2.64	1.89	1.67	3.31	2.65
	Number of visits Page 2	3.21	3.04	1.95	1.59	2.54	2.45
	Number of visits Page 3	3.17	3.05	2.59	2.29	2.86	2.68
Number of videos started		0.80^a*	0.91	1.00	1.05	0.91	0.99
Number of (re)starts of videos		0.87	1.05	1.08	1.17	0.98	1.12
	Number of starts Video 1	0.32	0.50	0.37	0.55	0.35	0.53
	Time spent on Video 1 (%)^b	16.64	11.71	15.56	12.78	16.03	12.29
	Number of starts Video 2	0.31	0.53	0.35	0.52	0.33	0.52
	Time spent on Video 2 (%)^b	32.50	17.69	29.61	16.39	30.81	16.92
	Number of starts Video 3	0.23	0.46	0.35	0.53	0.30	0.50
	Time spent on Video 3 (%)^b	19.45	9.86	21.98	10.10	21.10	9.98
Number of mode actions		4.87	3.01	-	-	-	-
Time until first mode action		18.47	14.62	-	-	-	-
		n	%			n	%
First mode chosen
	Text	139	76.0	-	-	-	-
	Visuals	25	13.7	-	-	-	-
	Video	19	10.4	-	-	-	-
Mode combinations
	All three modes	155	84.7	-	-	-	-
	Text and visuals	16	8.7	-	-	-	-
	Text and video	1	0.5	-	-	-	-
	Visuals and video	1	0.5	-	-	-	-
	Text-only	8	4.4	-	-	-	-
	Visuals-only	1	0.5	-	-	-	-
	Video-only	1	0.5	-	-	-	-

Notes. ^aMean differs significantly from the standardized website condition. ^bThe percentage of the video watched was calculated based on the first view. Page 2 = “My GIOCA-day”; Page 3 = “How to prepare”; Video 1 = “What does a GIOCA-day look like” (2:53 min); Video 2 = “Practical information for a hospital visit” (1:09 min); Video 3 = “How to prepare for the medical consultation” (2:37 min).

* p <.05. *** p <.001.

Table 3. Standardized direct and indirect effects of mode tailoring on website attitude and information recall (N = 363)

Path	Estimate	SE	95% CI	p-value
Direct paths
Condition (Mode tailoring) → Active control	.11*	.05	[0.01, 0.21]	.034
Active control → Perceived relevance	.61***	.04	[0.53, 0.67]	<.001
Active control → Website engagement	.22***	.05	[0.13, 0.32]	<.001
Active control → Cognitive load	−.51***	.04	[−0.59, −0.43]	<.001
Active control → Website attitude	.44***	.05	[0.34, 0.54]	<.001
Active control → Information recall	.08	.06	[−0.03, 0.19]	.170
Perceived relevance → Website attitude	.38***	.05	[0.29, 0.46]	<.001
Perceived relevance → Information recall	.00	.06	[−0.10, 0.11]	.952
Website engagement → Website attitude	.03	.04	[−0.04, 0.10]	.428
Website engagement → Information recall	.38**	.05	[0.28, 0.47]	.001
Cognitive load → Website attitude	−.04	.04	[−0.12, 0.05]	.411
Cognitive load → Information recall	−.26***	.04	[−0.34, −0.18]	<.001
Indirect paths
MT → AC → PR	.07*	.03	[0.01, 0.13]	.033
MT → AC → WE	.03*	.01	[0.00, 0.05]	.022
MT → AC → CL	−.06*	.03	[−0.11, −0.01]	.031
AC → PR/WE/CL → WA	.25***	.04	[0.18, 0.33]	<.001
AC → PR/WE/CL → IR	.22***	.04	[0.13, 0.31]	<.001
MT → AC → PR/WE/CL → WA	.08*	.04	[0.01, 0.15]	.033
MT → AC → PR/WE/CL → IR	.03*	.02	[0.01, 0.07]	.026
Total effects
AC → WA	.69**	.03	[.63,.75]	.001
AC → IR	.30***	.05	[.21,.38]	<.001
MT → WA	.08*	.04	[.01,.15]	.033
MT → IR	.03*	.02	[.01,.07]	.026

Note. Standardized estimates with bootstrapped standard errors and bias-corrected significance levels are presented. Abbreviations: MT = Mode tailoring. AC = Perceived active control. PR = Perceived relevance. WA = Website attitude. WE = Website engagement. CL = Cognitive load. IR = Information recall.

Figure 2. Hypothesized model. Model fit (N = 363): χ2 (9) = 18.11, p =.034, CMIN/DF = 2.012, TLI = 0.971, CFI = 0.987, RMSEA =.053, SRMR =.040, ECVI = 0.155. Website attitude: R² =.573; Recall of information: R² =.274. Estimates presented are standardized path estimates. Including the background variables as control (i.e., age, gender, education, cancer experience, Internet experience) in the model did not change the significance levels of the estimated paths.