Usability evaluation of the domain specific language for spatial simulation scenarios

Figures & data

Figure 1. The DSL3S meta-model.

Figure 2. The tools used to implement DSL3S.

Table 1. Summary of the tasks performed during the evaluation exercise

Phase	No.	Task
Tool set up	1	Install plug-ins
	2	Set up Java libraries
	3	Obtain spatial datasets
Project set up	4	Create project
	5	Set spatial datasets
	6	Create model
	7	Apply DSL3S profile
	8	Create simulation element
Basic prey	9	Create pasture element
	10	Create prey
	11	Add prey view
	12	Define Prey state
	13	Set prey grazing pasture
	14	Transform model to code and run
Refine prey	15	Set Prey replication
	16	Set prey perishing
	17	Set prey movement
	18	Transform model to code and run
Predator	19	Create predator element
	20	Add predator view
	21	Define predator state
	22	Set predator replication
	23	Set predator perishing
	24	Set predator movement
Animat interactions	25	Add interactions view
	26	Create Supply element
	27	Create Harvest element
	28	Associate Harvest with Supply
	29	Transform model to code and run
Inaccessible areas	30	Create Inaccessible Areas element
	31	Add inaccessible areas to prey view
	32	Modify prey movement
	33	Transform model to code and run
	34	Modify layer order (if needed)

Figure 3. Still from a run of the model developed during the evaluation sessions.

Figure 4. Age and gender profiles of the participants.

Figure 5. Distribution of blocking issues.

Figure 6. Exercise completion distribution.

Table 2. Summary statistics for the language questions

	$\nu$	$\mu$	$\sigma$	${\gamma }_1$
Number of concepts	3.00	3.32	1.00	0.01
Notation	3.00	3.58	0.69	0.68
Ease of learning	4.00	3.79	0.54	$-2.25$
Suitability	3.00	3.37	0.68	$-0.54$

Figure 7. Mosaic diagram for the language questions

Table 3. Summary statistics for the tool usability questions

	$\nu$	$\mu$	$\sigma$	${\gamma }_1$
DSL3S plug-ins	4.00	4.00	0.97	$-1.46$
Papyrus	4.00	3.69	1.02	$-0.84$
Code generation	5.00	4.44	0.86	$-1.43$
Whole	4.00	3.94	0.80	$-0.56$

Figure 8. Mosaic diagram for the tool usability questions.

Table 4. Summary statistics for the general approach questions

	$\nu$	$\mu$	$\sigma$	${\gamma }_1$
Lacking skills	3.00	3.47	0.91	0.29
Communication	4.00	3.63	0.76	$-0.78$
Standard language	3.00	3.21	0.86	0.13
Own development	3.00	3.21	0.79	$-0.35$

Figure 9. Mosaic diagrams for the general approach questions.

Table 5. ${\chi }^2$ test results for computer scientists vis-à-vis other backgrounds

		${\chi }^2$	df	p-value
Language	Number of concepts	2.713	4	0.6070
	Notation	2.096	2	0.3506
	Ease of learning	0.950	2	0.6219
	Suitability	0.560	2	0.7558
Tools	DSL3S plug-ins	3.420	3	0.3313
	Papyrus	1.940	3	0.5849
	Code generation	6.073	3	0.1081
	Whole	3.933	3	0.2688
Approach	Lacking skills	2.599	3	0.4577
	Communication	2.956	3	0.3985
	Standard language	12.480	3	0.0059
	Own development	1.738	2	0.4194

Table 6. ${\chi }^2$ test results for participants experienced with spatial simulation vis-à-vis those not experienced

		${\chi }^2$	df	p-value
Language	Number of concepts	4.414	4	0.3528
	Notation	1.318	2	0.5173
	Ease of learning	0.735	2	0.6924
	Suitability	0.855	2	0.6523
Tools	DSL3S plug-ins	4.535	3	0.2092
	Papyrus	0.662	3	0.8821
	Code generation	3.158	3	0.3679
	Whole	9.164	3	0.0272
Approach	Lacking skills	1.428	3	0.6991
	Communication	3.242	3	0.3558
	Standard language	7.719	3	0.0522
	Own development	1.042	2	0.5939

Figure 10. Results for computer scientists vs. others (left) and experienced vs. non experienced in spatial simulation (right) regarding the Standard language question.

Figure 11. Results for experienced and non experienced in spatial simulation regarding the question on the development process as a whole