3. Affective outcomes analysis
3.2 Moderator analysis on affective outcomes
The homogeneity statistic is significant on affective outcomes; therefore, it is necessary to examine the possible moderators. The researcher separated moderators into six categories: study characteristics, sample characteristics, research design characteristics, program characteristics, digital game characteristics, and digital game categories to perform the analysis.
3.2.1 Examination of study characteristics on affective outcomes
The researcher separated study characteristics into three moderators: subject matter, publication Type and publication year and then examined possible moderators (see Table 4.17).
(1) Subject matter
The researcher separated subject matter into language, math, nature and science, social study, health and PE, computer science, and others. Research results showed that except social study, each subject matter weighted mean ES is positive and 95% CI is significantly different from zero, meaning that in these subject matters, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because subject matter (QB = 72.87, p<.05) achieved significant, it meant that the moderator affects affective outcomes. Post hoc showed that the effects of language (ES=1.07) is significantly higher than other kind of subject matter;
nature and science (ES=.60) is significantly higher than math (ES=.14).
Language and nature and science Qw achieve significant level (p<.05), meaning
that unknown moderators still exit.
(2) Publication type
The researcher separated publication type into journal and proceedings.
Research results showed that only journal weighted mean ES is positive and 95% CI is significantly different from zero, meaning that only in journal type, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
(3) Publication year
The researcher separated publication year into 1981-1985, 1986-1990, 1991-1995, 1996-2000, 2001-2005, 2006-2010, and 2011-2015. However, most studies were published during 2006-2010 and 2011-2015. Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that articles published during 2006-2010 and 2011-2015 showed the result that the effect of applying DGBL on students’
affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Table 4.17
Note. *: significant at the .05 level; d+:
weighted mean effect size; k: number of articles; CI: confidence interval; #: d
+positive and 95% CI significantly different from zero ;(): post hoc.
3.2.2 Examination of sample characteristics on affective outcomes
The researcher separated sample characteristics into two moderators:
educational level and sample location, and then examined possible moderators (see Table 4.18).
(1) Educational level
The researcher separated educational level into elementary, junior high school, senior high school, and university. Research results showed that except unspecified, each educational level weighted mean ES is positive and 95% CI is
significantly different from zero, meaning that in these educational level, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because educational level (QB = 8.59, p >.05) is insignificant, it meant that educational level is not the moderator of affective outcomes.
(2) Sample location
The researcher separated sample location into Asia, Europe, Africa, South America, North America, and Oceania. Due to the reason that the data of Africa and Oceania are none (k=0), they would be excluded. Research results showed that except South America(k=2), each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that in Asia, Europe, and North America, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because sample location (QB = 42.92, p<.05) achieved significant level, it meant that sample location is the moderator of affective outcomes. Post hoc showed that Asia (ES=.74) is significantly better than Europe (ES=.22), South America (ES=.31) and North America (ES=.24). Sample location Qw is significant (p<.05), meaning that unknown moderators still exit.
Table 4.18
Note. *: significant at the .05 level; d+:
weighted mean effect size; k: number of articles; CI: confidence interval; #: d
+positive and 95% CI significantly different from zero; (): post hoc.
3.2.3 Examination of research design characteristics on affective outcomes
The researcher separated research design characteristics into three moderators:
instrumentation, instructor bias and experiment design and then examined possible moderators (see Table 4.19).
(1) Instrumentation
The researcher separated instrumentation into self-compiled test and standardized test. Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that in self-compiled test and standardized test, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying
Non-DGBL on students’ affective outcomes.
Because instrumentation (QB = 5.2, p<.05) achieved significant level, it meant that instrumentation is the moderator of affective outcomes. Standardized test (ES=.52) is significantly better than Self-compiled test (ES=.33).
(2) Instructor bias
The researcher separated instructor bias into the same teacher (ST); different teacher (DT); mixture (MI); control group has no teacher involved and game as teacher in the experimental group (CNT>); control group has teacher involved and game as teacher in the experimental group (CT>); unspecified.
MI and unspecified data is none (k=0).
Research results showed that except CT and GT, each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that in these instructor biases, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’
affective outcomes.
Because instructor bias (QB = 51.13, p<.05) achieved significant level, it meant that instructor bias is the moderator of affective outcomes. Post hoc showed that ST (ES=.94) is significantly higher than CNT and GT (ES=.46) and DT (ES=.23); CNT and GT is significantly higher than DT. Except CTandGT, each instructor bias Qw are significant (p<.05), meaning that unknown moderators still exit.
(3) Experiment design
The researcher separated experiment design into one-group pretest posttest design (OPP), pretest-posttest control group design (PPC), posttest-only control group design (POC), the nonequivalent pretest-posttest design (NPP), the nonequivalent posttest only design (NPO) and unspecified. Research results showed that except POC (k=3), each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because experiment design (QB = 13.80, p<.05) achieved significant level, it meant that experiment design is the moderator of affective outcomes. Post hoc showed that OPP (ES=.60) and NPP (ES=.50) are significantly better than PPC (ES=.28). Except POC, each experiment designs Qw are significant (p<.05), meaning that unknown moderators still exit.
Table 4.19
Note. *: significant at the .05 level; d+:
weighted mean effect size; k: number of articles; CI: confidence interval; #: d
+positive and 95% CI significantly different from zero; (): post hoc.
3.2.4 Examination of program characteristics on affective outcomes
The researcher separated program characteristics into four moderators: duration of treatment, purpose of treatment, group size in experiment group and strategy involved and then examined possible moderators (see Table 4.20).
(1) Duration of treatment
The researcher separated duration of treatment into less than 7 days, 8-30 days, above 31 days, and unspecified. Research results showed that except
unspecified, each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that the effect of applying DGBL on students’
affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because duration of treatment (QB = 25.55, p<.05) achieved significant level, it meant that duration of treatment is the moderator of affective outcomes. Post hoc showed that less than 7 days (ES=.52) and 8-30 days (ES=.46) are significantly better than above 31 days (ES=.20). Except unspecified, each experiment designs Qw are significant (p<.05), meaning that unknown moderators still exit.
(2) Purpose of treatment
The researcher separated purpose of treatment into superior, complement, and replace. Research results showed that except replace, each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because purpose of treatment (QB = 7.68, p<.05) achieved significant level, it meant that purpose of treatment is the moderator of affective outcomes. Post hoc showed that superior (ES=.40) is significantly higher than complement (ES=.20). Except replace, superior and complement Qw are significant (p<.05), meaning that unknown moderators still exit.
(3) Group size in experiment group
The researcher separated group size in experiment group into individual and group. Research results showed that each weighted mean ES is positive and 95% CI is significantly different from zero, meaning that in both individual and group, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because group size in experiment group (QB = .06, p>.05) is not significant, it meant that group size in experiment group is not the moderator of affective outcomes. Individual Qw are significant (p<.05), meaning that unknown moderators still exit.
(4) Strategy involved
The researcher separated strategy involved into both does not have strategy (BNS); both have strategies (BHS); control group has strategy but experimental group doesn’t apply any strategy (CHEN); experimental group has strategy but control group has no strategy (EHCN). The data of BHS are none. Research results showed that except EHCN (k=2), BNS and CHEN weighted mean ES are positive and 95% CI are significantly different from zero, meaning that if both groups have no strategy involved or if control group has strategy but experimental group doesn’t apply any strategy, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because strategy involved (QB = 49.84, p<.05) achieved significant level, it
meant that the strategy involved is the moderator of affective outcomes. Though post hoc showed that CHEN (ES=1.71) is significantly better than BNS (ES
= .32), the conclusion is not confirmed because the article numbers are seriously unbalanced (CHEN k=3). Each strategy involved Qw is significant, meaning that unknown moderators still exit.
Note. *: significant at the .05 level; d+:
weighted mean effect size; k: number of articles; CI: confidence
interval; #: d
+positive and 95% CI significantly different from zero; (): post hoc.
3.2.5 Examination of digital game categories on affective outcomes
The researcher separated digital game categories into action games, adventure games, fighting games, puzzle games, role-playing games, simulation games, sports games, strategy games, and unspecified. Due to the reason that the data of action games, adventure games, fighting games and sports games are few; these game categories will not be discussed. Research results showed that puzzle games, role-playing games, simulation games, and strategy games weighted mean ES are positive and 95% CI are significantly different from zero, meaning that the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes (see Table 4.21).
Because digital game categories (QB = 5.08, p>.05) is insignificant, it meant that digital game categories is not the moderator of affective outcomes. Puzzle games, role-playing games and simulation games Qw are significant, meaning that unknown moderators still exit.
Table 4.21
Results of Digital Game Categories Examination
Q
B Kd
+ 95%CIQ
wDigital game categories 5.08 31
4.Puzzle games 16 .32# [.24, .40] 107.39*
5. Role-playing games 4 .45# [.08, .82] 56.98*
6.Simulation games 6 .52# [.28, .77] 28.12*
8.Strategy game 5 .33# [.14, .53] 6.36
Note. *: significant at the .05 level; d+:
weighted mean effect size; k: number of articles; CI: confidence
interval; #: d
+positive and 95% CI significantly different from zero;(): post hoc.
3.2.6 Examination of digital game characteristics on affective outcomes
The researcher separated digital game characteristics into intrinsic vs. extrinsic, tightly linked vs. loosely linked, hard-wired vs. engines and templates or shells, reflective vs. action, synchronous vs. asynchronous, single-player vs.
multiplayer, session-based games vs. persistent-state, video-based vs.
animation-based and narrative-based vs. reflex-based (see Table 4.22).
(1) Intrinsic vs. Extrinsic
Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that in intrinsic or extrinsic, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL. Because intrinsic vs. extrinsic (QB
= .49, p>.05) is insignificant, it meant that intrinsic vs. extrinsic is not the moderator of affective outcomes.
(2) Tightly linked vs. Loosely linked
Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that no matter in tightly linked or loosely linked, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes.
Because tightly linked vs. loosely linked (QB = 12.18, p<.05) achieved significant level, it meant that tightly linked vs. loosely linked is the moderator of affective outcomes, and that loosely linked (ES=.83) is significantly higher than tightly linked (ES=.33).
(3) Hard-wired vs. Engines and templates or shells
Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that in hard-wired or engines and templates or shells, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes. Because hard-wired vs. engines and templates or shells (QB = .14, p>.05) is insignificant, it meant that hard-wired vs. engines and templates or shells is not the moderator of affective outcomes.
(4) Reflective vs. Action
All studies are Reflective. Therefore, comparison is impossible.
(5) Synchronous vs. Asynchronous
Research results showed that both synchronous and asynchronous weighted mean ES are positive and 95% CI are significantly different from zero, meaning that no matter in synchronous or asynchronous, the effect of applying DGBL on students’ affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes. Because synchronous vs.
asynchronous (QB = .20, p>.05) is insignificant, it meant that synchronous vs.
asynchronous is not the moderator of affective outcomes.
(6) Single-player vs. Multiplayer
Research results showed that both single-player and multiplayer weighted mean ES are positive and 95% CI are significantly different from zero, meaning that no matter single-player or multiplayer, the effect of applying DGBL on students’
affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes. Because single-player vs. multiplayer (QB = 1.51, p>.05) is insignificant, it meant that single-player vs. multiplayer is not the moderator of affective outcomes.
(7) Session-based games vs. Persistent-state games
Research results showed that both weighted mean ES are positive and 95% CI are significantly different from zero, meaning that no matter in session-based games or persistent-state games, the effect of applying DGBL on students’
affective outcomes is significantly better than the effect of applying Non-DGBL on students’ affective outcomes. Because session-based games vs.
persistent-state (QB =3.85, p<.05) achieved significant level, it meant that session-based games vs. persistent-state games is the moderator of affective outcomes, and that session-based games (ES=.43) is significantly higher than persistent-state games (ES=.31).
Table 4.22
Results of Digital Game Characteristic Examination
Q
B K d+ 95%CIQ
wDigital game characteristics
Intrinsic vs. Extrinsic 0.49 33
1.Intrinsic 8 .31# [.13, .49] 10.20
2.Extrinsic 25 .36# [.30, .43] 194.59*
Tightly Linked vs. Loosely Linked 12.18* 33
1.Tightly Linked 29 .33# [.27, .40] 129.68*
2.Loosely Linked 4 .83# [.39, 1.27] 63.40*
Hard-Wired vs. Engines and
Templates or Shells 0.14 32
1.Hard-Wired 11 .37# [.26, .47] 19.38*
2.Engines and Templates or Shells 21 .34# [.26, .43] 185.58*
Synchronous vs. Asynchronous 0.20 33
1. Synchronous 25 .36# [.29, .43] 114.34*
2. Asynchronous 8 .33# [.17, .49] 90.73*
Single-Player vs. Multiplayer 1.51 33
1.Single-Player 27 .37# [.31, .44] 183.47*
2.Multiplayer 6 .28# [.10, .46] 20.29*
Session-Based vs. Persistent-State 3.85* 33
1. Session-Based Games 21 .43# [.33, .52] 130.07*
2. Persistent-State Games 12 .31# [.22, .39] 71.35*
Note. *: significant at the .05 level; d+: