Content uploaded by Scott Frankowski
Author content
All content in this area was uploaded by Scott Frankowski on Sep 25, 2020
Content may be subject to copyright.
_______________________________
Author info: Correspondence should be sent to: James C. Lue, Department of
Psychology, Midwestern State University, 3410 Taft Blvd, Wichita Falls, TX
76308, USA. Email: james.c.lue@gmail.com
North American Journal of Psychology, 2020, Vol. 22, No. 3, 499-520.
NAJP
High Need for Drama Individuals May Interpret
Their Manipulative and Impulsive Behaviors as
Self-Efficacy
James C. Lue and Scott D. Frankowski
Midwestern State University
This research explores the relationships between need for drama (NFD),
self-efficacy, and self-esteem. People who have dramatic tendencies tend
to be manipulative, impulsively outspoken, and believe they are
consistently the victim of circumstance. High NFD individuals tend to
have low self-esteem. Self-esteem and self-efficacy strongly correlate. In
the present research, we tested the hypothesis that higher levels of NFD
would correlate with lower levels of self-efficacy. High NFD individuals’
cycles of maladaptive behaviors may be over-compensation for negative
self-evaluations and limited self-efficacy. Study 1 showed that NFD did
not correlate with self-efficacy in a student and an M-Turk sample.
Exploratory analyses did, however, show a quadratic U-shaped
relationship in the M-Turk sample such that self-efficacy decreased as
NFD increased but self-efficacy increased at the high end of the NFD
scale. Study 2 showed a weak negative linear association between NFD
and self-efficacy, but we replicated the strong quadratic relationship
found in Study 1. There was also a strong inverted U-shaped relationship
between NFD and stress. High NFD individuals may interpret their
impulsive and manipulative tendencies as self-efficacy, which may
contribute to the cycle of drama in their lives, but which also may buffer
them from stress. A secondary research aim was discussed, as were future
directions.
Keywords: need for drama, dramatic personality, manipulative behavior,
self-efficacy, self-esteem
Individuals with dramatic dispositions are often difficult to deal with.
Their lives are marked by social disorder and contrived emergencies.
Frankowski et al. (2016) developed a measure of dramatic behavioral
intentions to assess the extent to which people impulsively manipulate
others from a position of perceived persistent victimization. This Need
for Drama (NFD) construct is represented by three intercorrelated
factors: interpersonal manipulation (IPM), impulsive outspokenness (IO),
and persistent perceived victimization (PPV). The NFD scale positively
500 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
correlates with the dark triad (see Paulhus & Williams, 2002 for
overview on the dark triad), attitudes toward gossip, and neuroticism
(Frankowski et al., 2016). The NFD measure negatively correlates with
conscientiousness, agreeableness, and self-esteem. The lack of self-
esteem among high-NFD individuals suggests that these individuals’
manipulative and impulsive behaviors may be an attempt to compensate
for their negative self-evaluations. Such a cycle of impulsive and
manipulative behaviors may lead to lower levels of self-efficacy, which
may contribute to feelings of persistent perceived victimization, and in
turn fuel the cycle of dramatic behaviors.
Our primary objective for this research was to better understand how
NFD is related to self-efficacy. Self-efficacy is a person’s evaluation of
their aptitude to perform an activity, attain an objective, or overcome a
barrier (Bandura, 1994). Ewen (2014) clarified that “perceived self-
efficacy is not simply a matter of ‘talking a good game’; it involves
fundamental beliefs about yourself” (p. 346). These beliefs determine
how individuals think, feel, behave, and self-motivate. A robust sense of
self-efficacy enhances achievement and personal welfare (Ewen, 2014).
Individuals with faith in their competence approach difficult tasks as
challenges to conquer instead of threats to dodge. They create
challenging objectives and maintain firm dedication to these objectives.
However, individuals who harbor self-doubt in their skills avoid difficult
activities, which they consider to be personal threats. Such individuals
have low ambitions and feeble commitment to the objectives they choose
to undertake (Ewen, 2014). Not surprisingly, self-efficacy is very
strongly correlated with self-esteem (see meta-analysis by Judge & Bono,
2001).
A negative relationship between NFD and self-esteem was
established by Frankowski et al. (2016) but has not yet been explored
further. Self-esteem is the extent to which one prizes, values, approves,
or likes oneself (Blascovich & Tomaka, 1991). Self-esteem is usually
considered to be the evaluative component of self-concept, an
individual’s broad representation of self that contains cognitive,
emotional, and behavioral components.
The strong relationship between self-esteem and self-efficacy (Judge
& Bono, 2001) suggests that those high in NFD, in addition to having
lower levels of self-esteem, will likely also have lower levels of self-
efficacy. The manipulative and impulsive behaviors, and the feelings of
persistent perceived victimization may be a reaction to lower levels of
self-efficacy and self-esteem in high NFD individuals. Additionally,
those who have low levels of self-efficacy tend to score high on
neuroticism and low on conscientiousness (Hartman & Betz, 2007). This
personality profile of low self-efficacy matches NFD’s correlates to
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 501
neuroticism and conscientiousness (Frankowski et al., 2016). Thus, we
expected that high NFD individuals would also have lower self-efficacy.
Frankowski et al. (2016) found that the relationship between self-esteem
and the persistent perceived victimization factor of NFD was particularly
strong. Thus, we expected that this factor of NFD would correlate more
strongly with self-efficacy than the interpersonal manipulation and social
impulsivity factors.
With any measure of traits that may be considered socially off-
putting, research participants may try to present themselves in a better
light when self-reporting (Furr, 2010). Individuals may underreport their
manipulative and dramatic tendencies to make themselves look better.
Socially desirable responding can compromise the validity of research.
Furthermore, social desirability also occurs outside of research
environments. High-stakes situations, such as selecting job applicants,
often drives up the motivation of “faking good” on personality
assessments (Fahey, 2018). In relation to the primary aim of this
research, if the NFD measure positively correlates with social
desirability, it would be difficult to determine whether the relationship
between NFD and self-efficacy is influenced by individuals trying to
present themselves as more competent than they actually are. Conversely,
high NFD individuals may be less likely to self-monitor due to their
impulsive tendencies, in which case we would expect a negative
correlation between NFD and social desirability. As a secondary aim of
this research, we tested the relationship between NFD and social
desirability to gauge the construct validity of the measure.
Study 1 Hypotheses
1. We hypothesized that NFD would negatively correlate with self-
efficacy, and that the persistent perceived victimization factor
would correlate more strongly with self-efficacy than the
interpersonal manipulation or social impulsivity factors.
2. We hypothesized that NFD would not correlate, or would
correlate negatively, with social desirability.
Hypotheses were registered prior to data collection on the Open
Science Framework (OSF). Additional analyses not pre-registered are
presented as exploratory. All studies conducted are reported, as are all
measures within the studies, and any participant exclusions. Data, syntax,
and analyses are available on the OSF at https://osf.io/y3pzw/.
502 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
STUDY 1
Method
For Study 1, we collected data from a sample of university students
and an M-Turk sample. We pre-registered a plan to collect a sample size
of at least 200 for a single sample to detect moderately small correlations.
But because of demographic differences, namely age, which has been
shown to negatively correlate with NFD (Frankowski et al., 2016), we
analyzed the student (Study 1a) and M-Turk (Study 1b) samples
separately.
Study 1a Participants Data were collected from 210 participants at a
small public university in the United States. There were 154 women and
55 men and one participant who did not identify their gender.
Participants’ mean age was 20.27 ± 3.84 years. Eighty-two (39.23%)
were White, 50 (23.81%) were Black or African American, 44 (20.95%)
were Latino or Hispanic, 14 identified as multi-racial (6.70%), 12
(5.74%) were Asian or Pacific Islanders, 5 (2.39%) were Native
Americans, and two participants (0.96%) indicated “Other.”
Study 1b Participants Data were collected from 140 participants using
the M-Turk platform with location restricted to the United States (no
other restrictions per Robinson et al., 2019). There were 61 women
(43.88%) and 78 men (56.12%). Participants’ mean age was 33.73 ± 9.48
years. The sample was mostly White (n = 115; 82.14%) but included
eight (5.71%) who were Black or African American, six (4.29%) who
were Asian or Pacific Islander, four (2.86%) who were Latino or
Hispanic, four (2.86%) who were Native American, and three (2.14%)
who identified as multiracial.
Materials We measured dramatic personality with the 12-item NFD scale
(Frankowski et al., 2016). The scale measures dramatic behavioral
intentions and self-evaluations on three factors: (1) interpersonal
manipulation (four items; e.g. “Sometimes I play people against each
other to get what I want.”); (2) impulsive outspokenness (three items; e.g.
“I wait before speaking my mind.” – reverse coded); and (3) persistent
perceived victimization (five items; e.g. “A lot of people have wronged
me.”). Participants responded on a 7-point Likert-type scale (Strongly
disagree to Strongly agree). An average of each subscale was obtained,
and the subscales were averaged to score the full measure.
We measured self-efficacy with the 10-item Generalized Self-Efficacy
scale (Schwarzer & Jerusalem, 1995). This measure gauges the general
belief in oneself to solve problems and reach goals and is one of the most
popular scales to measure self-efficacy (e.g. “It is easy for me to stick to
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 503
my aims and accomplish my goals.”). Response options for each
statement were Not at all true, Hardly true, Moderately true, and Exactly
true. The score was calculated by averaging the items. Higher scores
suggest a greater level of self-efficacy.
We measured social desirability using the 13-item short form of the
Marlowe-Crowne Social Desirability Scale (Reynolds, 1982). The items
were answered with True or False, with items being coded with regard to
whether it is a socially desirable answer or not. An average of responses
was computed ranging from 0 to 1, which is a proportion of how many
questions were answered in a socially desirable manner.
Procedure Data were collected online with Google Forms. After
participants read and agreed to the informed consent, they were
instructed to begin the survey. Responses were optional for all items. We
randomized the order of the scales, which appeared as NFD, social
desirability, and self-efficacy, respectively. After the last scale,
volunteers provided demographic information. Participants were
debriefed and thanked for their time. We disclosed that the purpose of
our research is to evaluate how dramatic traits are associated with both
social desirability and self-efficacy.
Prior to analyses, the data were checked for missing values,
skewness, and kurtosis. Participants missed six items in Study 1a
(0.08%) and 20 items in Study1b (0.40%). Missing values were imputed
using a single imputation procedure in SAS. Outliers can affect
relationships among variables (Burdenski, 2000), but we did not
preregister a plan to identify or exclude outliers. Examining bivariate
plots and frequency tables, however, did not reveal any obvious outliers.
The skew and kurtosis were within acceptable ranges (< 2 and < 7
respectively; Curran et al., 1996; Table 1).
Results
Descriptive and normality statistics for both samples are presented in
Table 1. Reliability and correlations are in Table 2. Neither the full NFD
measure nor the subscales correlated with self-efficacy, contrary to
hypotheses. As an exploratory analysis, we also checked whether a
curvilinear model would fit the data. Perhaps as NFD tendencies
increase, the individuals experience less self-efficacy until at high levels
of NFD, they experience an increase in self-efficacy (i.e. a quadratic
relationship). Or, it could be the case that deterrence to self-efficacy as a
function of personality operates similar to affective deterrence and group-
identity deterrence (i.e. a cubic relationship; see e.g. Brehm, 1999;
Pantaleo et al., 2014) – in this model, self-efficacy would decrease to a
point as NFD increases, increase in the mid-to-high range of NFD, and
504 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
decrease again at the highest levels of NFD. In the student sample,
neither of these possibilities appear to be the case, as neither a quadratic
model (F(2, 207) = 0.22, p =.81) nor a cubic model (F(3, 206) = 0.37, p
=.77) explained significant variance in the data. In the M-Turk sample,
however, a quadratic model explained a large amount of the variance,
F(2, 137) = 7.16, p = .001, R
2
= .09 (Figure 1) showing that as NFD
increased, self-efficacy decreased, but increased at the highest levels of
NFD. A cubic model did not explain a significant amount of variance
beyond the quadratic model in the M-Turk sample, F(1, 137) = 0.27, p =
.60.
There was a strong negative correlation between NFD and social
desirability in the student sample and a weak-to-moderate negative
correlation in the M-Turk sample. Notably, the full NFD measure had
poor reliability with this student sample. The factor intercorrelations
were low and thus the correlates to each individual factor should be
examined rather than the full measure in the student sample. There was
good alpha reliability and strong intercorrelations between factors of the
NFD measure in the M-Turk sample.
Table 1 Descriptive and Univariate Normality Statistics
Study 1a (N = 210)
Student Sample
Study 1b (N = 140)
M-Turk Sample
Mean
(SD)
Skew Kurtosis Mean
(SD)
Skew Kurtosis
NFD (Full
measure)
3.54
(1.01)
0.08 -0.48 3.46
(1.48)
0.16 -0.88
IPM
subscale
2.87
(1.36)
0.67 -0.22 3.34
(1.89)
0.28 -1.28
IO subscale 3.64
(1.39)
0.13 -0.64 3.35
(1.36)
-0.05 -0.60
PPV
subscale
4.10
(1.47)
0.03 -0.72 3.68
(1.75)
0.09 -1.20
Social
desirability
0.47
(0.24)
-0.03 -0.85 0.45
(0.25)
0.13 -0.44
Self-efficacy 3.07
(0.46)
-0.07 -0.19 3.05
(0.59)
-0.96 1.92
Note. IPM = Interpersonal manipulation, IO = Impulsive outspokenness, PPV = Persistent
perceived victimization.
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 505
Table 2 Alpha Values and Correlations among Variables in Studies 1a
and 1b
1 2 3 4 5 6
1. Full NFD
measure
.53
.87
2. IPM
subscale
-- .80
.92
3. IO subscale -- .22
b
,
.66
c
.59
.68
4. PPV
subscale
-- .37
c
.68
.22
b
.70
c
.81
.91
5. Self-efficacy .00
-.01
-.05
.12
.01
-.05
.01
-.12
.82
.89
6. Social
desirability
-.46
c
-.23
b
-.45
c
-.16
-.22
b
-.22
b
-.33
c
-.24
b
.25
c
.11
.77
.76
Note.
a
p < .05,
b
p < .01,
c
p < .001. The top coefficient is from the student sample (Study 1a);
the bottom coefficient is from the M-Turk sample (Study 1b). Cronbach alpha coefficients
are listed on the diagonal. The alpha for the full NFD measure was computed with the three
subscales.
Figure 1 Quadratic Relationship between NFD and Self-efficacy in M-
Turk Sample
Discussion
We hypothesized that NFD would correlate negatively with self-
efficacy, particularly the persistent perceived victimization factor of
NFD. Contrary to this hypothesis, neither NFD nor its factors correlated
with self-efficacy. There were, however, surprising results when testing
506 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
exploratory non-linear models. In the M-Turk sample, there was a U-
shaped curve such that self-efficacy decreased in the mid-range of NFD
but rose at the high end of the measure. There was, however, no
curvilinear relationship between NFD and self-efficacy among the
student sample. The samples did not differ on their mean levels of self-
efficacy (Table 1). There were substantial demographic differences
between these two samples that may account for differences in how NFD
and self-efficacy are associated. The student sample was younger, more
racially and ethnically diverse, and have all taken steps to pursue a four-
year university degree. That they are all pursuing higher education, and
most are in the idealistic emerging adulthood stage of development
(Arnett, 2006) may contribute to their levels of self-efficacy remaining
consistent across low and high levels of NFD.
Whereas there is a U-shaped relationship between self-efficacy and
NFD that may be dependent on demographic characteristics, there is a
strong linear association between self-esteem and self-efficacy (Judge &
Bono, 2001) and between NFD and self-esteem (Frankowski et al.,
2016). It could be the case that NFD moderates the effect of self-esteem
predicting self-efficacy. Those at the low end of NFD may experience
increased self-efficacy, as they have more self-esteem, in-line with the
strong ρ = .85 relationship found in Judge and Bono’s (2001) meta-
analysis. Those at the high end of NFD, despite lower levels of self-
esteem, may have normal levels of self-efficacy that are moderated by
NFD personality characteristics. We tested this moderating hypothesis in
Study 2.
In Study 1, we found that NFD negatively correlated with social
desirability. In the student sample, this was a strong relationship, and in
the M-Turk sample, this was a moderate relationship. Notably, the
relationship between NFD and social desirability is similar to the
relationship between neuroticism (and related traits: anxiety, hostility,
depression, impulsivity) and social desirability (McCrae & Costa, 1986).
High NFD individuals are less concerned with self-presentation, despite a
greater self-focus (Frankowski et al., 2016). The correlation between
NFD and social desirability also shares similarity to social desirability
having a negative relationship with Machiavellianism and psychopathy
(Kowalski et al., 2018). The difference in the strength of this relationship
between student and M-Turk samples may indicate that the relationship
between NFD and social desirability weakens with age. As people
transition from college to the working world, they may learn to self-
monitor and better regulate behaviors. Causing excessive chaos with co-
workers and managers can affect one’s employment status, so older high
NFD individuals may have learned to exercise restraint compared to their
younger high NFD counterparts.
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 507
STUDY 2
For Study 2, we hypothesized that NFD would moderate the
relationship between self-esteem and self-efficacy. Self-esteem and self-
efficacy may be significantly less correlated at the high end of the NFD
measure compared to lower levels of the measure. We also included
qualitative questions in this study in which participants recalled a time
when they completed an important task, and their emotions associated
with completing that task. Although we tested linear and curvilinear
relationships in Study 1, quantitative measures do not always reveal
complex relationships among constructs. Furthermore, a qualitative
analysis can inform quantitative findings and spur future research
(Heydarian, 2016). For this qualitative analysis, we hypothesized that
high NFD individuals would report fewer positive emotions and more
negative emotions when recalling completing an important task. In line
with the results of Study 1, we also hypothesized that NFD would not
significantly correlate with rater-coded self-efficacy when participants
recalled how taking on an important task influenced how they took on
other tasks. High NFD individuals may demonstrate similar levels of
self-efficacy as anyone else but may have more negative emotional
experiences associated with their self-efficacy. Quantitative hypotheses
were pre-registered on the Open Science Framework, and qualitative
hypotheses were registered as we began analyses and iteratively revisited
our coding (https://osf.io/y3pzw).
Method
Participants There were 146 M-Turk participants who completed Study
2. We decided our approximate sample size a priori and pre-registered
our intention to collect data from at least 130 participants to detect
moderately small correlations and a moderately small interaction term in
our moderation test. Three participants were excluded because they had
IP addresses that traced to outside of the United States. Thus, there were
143 participants included in analyses. There were 62 women (43.33%)
and 80 men (55.94%), and one participant identified as non-binary.
Participants’ average age was 33.73 ± 9.48 years. The sample was mostly
White (n = 99; 69.23%), but included 22 (15.38%) who were Asian,
seven (4.90%) who were Black or African American, seven (4.90%) who
were Latino or Hispanic, one American Indian (0.70%), and seven
(4.90%) listed other or multiracial.
Materials We used the same NFD and self-efficacy scales that we used in
Study 1. We measured self-esteem with the 10-item Rosenberg Self-
Esteem scale (Rosenberg, 1965). This scale measures global self-worth
by computing both positive and negative feelings about the self and is
508 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
one of the most widely used scales to measure self-esteem. Responses
were reported on a four-point scale (Strongly disagree to Strongly agree
with no unsure option). An average of the items was calculated. Higher
scores suggest a greater level of self-esteem.
In the qualitative section, participants were asked to consider a time
they attempted to complete a very important task in their life. We asked
three qualitative questions about the task and gave two additional
questions with Likert-type scale ratings that evaluated their perceptions
about the task. The three open-ended questions were: (1) “What was the
task?,” (2) “How did you feel when you were taking on the task? (e.g.
What emotions did you feel while taking on the task?),” and (3) “How
did completing this important task influence how you approach other
important tasks in your life?” After the open-ended response questions,
participants were asked to rate if they believed they completed the task
successfully (“Thinking of this task, to what extent do you believe you
completed the task successfully?” 1 = Definitely did not complete
successfully, 6 = Definitely completed successfully), and “Thinking
back, how stressed were you when taking on the task?” (1 = Not stressed
at all, 5 = Majorly stressed). These two Likert-type questions were
exploratory, and we did not have a priori predictions about them.
Procedure Data were collected with Qualtrics. Responses were optional
for all items. The order of the measures was randomized as were the
order of items within each measure. Randomizing the order of measures
and items eliminates systematic presentation order effects, with the trade-
off of introducing random error. After the last scale, participants
completed the qualitative questions and the two Likert-type questions
related to the qualitative questions. Finally, participants provided
demographic information. Participants were debriefed and thanked for
their time. We disclosed that the purpose of our research is to evaluate
how dramatic traits are associated with both self-esteem and self-
efficacy.
Prior to analyses, the data were checked for missing values,
skewness, and kurtosis. There were six missing values (0.1%) that were
imputed with a single imputation in SAS. The skew and kurtosis were
within acceptable ranges (< 2 and < 7 respectively; Curran et al., 1996).
Results
Descriptive and normality statistics are presented in Table 3.
Correlations and scale reliability are presented in Table 4. Notably, in
this sample there was a weak negative correlation between the full NFD
measure and self-efficacy. There was a moderate negative relationship
between the PPV factor of NFD and self-efficacy. This finding differs
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 509
from Study 1 in which there were non-significant relationships between
NFD and self-efficacy in both samples.
Table 3 Descriptive and Normality Statistics for Study 2
Study 2 (N = 143)
M-Turk Sample
Mean (SD) Skew Kurtosis
NFD (Full measure) 3.39 (1.15) 0.59 -0.75
IPM subscale 3.01 (1.61) 0.41 -0.86
IO subscale 3.45 (1.19) -0.11 -0.39
PPV subscale 3.70 (1.39) -0.35 0.11
Self-efficacy 3.03 (0.51) -0.43 1.62
Self-esteem 2.94 (0.65) -0.22 -0.25
Note. IPM = Interpersonal manipulation, IO = Impulsive outspokenness, PPV = Persistent
perceived victimization.
Table 4 Correlations and Alpha Values of Study 2 Variables
1 2 3 4 5 6
1. Full NFD
measure
.76
2. IPM
subscale
-- .90
3. IO
subscale
-- .56
c
. 64
4. PPV
subscale
-- .57
c
. 41
c
.84
5. Self-
efficacy
-.18
a
-.04 .05 -.36
c
.90
6. Self-
esteem
-.41
c
-.30
c
-.24
b
-.46
c
.72
c
.90
Note.
a
p < .05,
b
p < .01,
c
p < .001. Cronbach alpha coefficients are listed on the diagonal.
Moderation Analysis The moderating effect of NFD on the relationship
between self-esteem and self-efficacy was tested and probed using
510 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
Preacher et al.’s (2006) online interaction tool at www.quantpsy.org. We
conducted a regression analysis to obtain the coefficients and the
asymptotic covariance matrix to plug into the quantpsy tool (using
heteroscedastic method 3 per Hayes, 2018). Contrary to our hypothesis,
NFD did not moderate the effect of self-esteem predicting self-efficacy,
b
interaction
= -0.03, t(139) = -0.39, p = .70 (Table 5). Probing at both the
low and high ends of NFD (± 1 SD) self-esteem strongly predicted self-
efficacy: -1 SD of NFD: b =0.64, t(139) = 8.51, p < .001; +1 SD of NFD:
b = 0.56, t(139) = 3.09, p = .002 (Figure 2). Interestingly, there was a
marginal positive effect of NFD predicting self-efficacy when controlling
for self-esteem and the interaction term.
Table 5 Study 2 Moderation Analysis
b (β) t p
Variable
Self-esteem 0.60 (0.76) 6.66 <.001
Need for Drama 0.06 (0.13) 1.75 .08
Self-esteem x Need for Drama -0.04 (-0.05) -0.39 .70
Note. The outcome variable is self-efficacy. Unstandardized coefficients are reported with
standardized coefficients in parentheses. Heteroscedastic standard errors were used for
inferential tests. The full model was significant, F(3, 139) = 54.14, p < .001, R
2
= .54.
Figure 2 Moderation Analyses of Testing NFD Moderating the
Relationship between Self-esteem and Self-efficacy
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 511
Exploratory Analysis As in Study 1, we investigated possible non-linear
relationships between NFD and self-efficacy in Study 2. With this
sample, we replicated the strong quadratic relationship between NFD and
self-efficacy, F(2, 140) = 10.17, p < .001, R
2
= .13 (Figure 3). A cubic
model did not explain additional variance beyond the quadratic model,
F(1, 139) = 0.01, p = .90. Because we found this strong quadratic
relationship between NFD and self-efficacy, we also tested whether a
curvilinear relationship would explain additional variance beyond the
strong linear relationship between NFD and self-esteem (correlation in
Table 4). A quadratic term added to the linear model did not explain
additional variance, F(1, 140) = 0.49, p = .48.
Figure 3 Quadratic Relationship between NFD & Self-efficacy in Study 2
Qualitative Analysis Codeable qualitative responses were provided from
113 participants. After an initial read through of the responses, the
authors discussed categories to code, and subsequently began coding.
First, the authors and another researcher coded the category of the
important task (item: “What was the important task?”). These category
codings were submitted to a judge (second author), who made decisions
where there was disagreement among coders. Table 6 shows an overview
of the counts in each coding category. As an exploratory analysis, we
examined the correlations between these categories and the quantitative
measures. None of the quantitative measures correlated significantly with
the category codings.
512 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
Table 6 Frequencies of Coded Qualitative Response Categories when
Asked to Name an Important Task they had Completed
Category Count
Self-accomplishment 43
Work 39
School 32
Interpersonal 16
Self-care 11
Other 2
Note. n = 113. Not all participants in Study 2 provided codeable responses. Response could
be coded into more than one category. An example of self-accomplishment was “Achieve
audit goals in my working group, a record number in the company,” which was also coded
into the work category. An example of a school coding was “Graduating college with high
honors,” which was also coded into the self-accomplishment category. A response coded as
interpersonal was “Helped my son overcome an addiction.” A response coded as self-care
was “I got two dental implants and a third tooth extracted.”
Next, the authors coded the responses to the second qualitative
question (“What emotions did you feel while taking on the task?”) by
counting the number of positive and negative emotions the participant
listed. The between rater correlation for positive emotions was r = .90
and for negative emotions was r = .93, thus we calculated the average
between the two coders. On average, participants listed 0.77 positive
emotions (SD = 0.88) and 1.32 negative emotions (SD = 1.01). Some
examples of positively coded emotions were hopeful, satisfied, and
excited. Some examples of negatively coded emotions were stressed,
nervous, and anxious. Need for drama did not significantly correlate with
positive, r = -.09, p = .32, nor negative emotions, r = -.03, p = .68.
Exploring the sub-factors of NFD, the IPM subscale correlated
moderately with negative emotions, r = -.21, p = .02; the other subscales
did not significantly correlate with positive nor negative emotions.
For the third question (“How did completing this important task
influence how you approach other important tasks in your life?”), we
originally planned to code responses as approach (e.g. gave the
confidence to take on new tasks) or avoid (e.g. avoid taking on new
tasks; Green, 1992). Responses, however, were not consistently codeable
by the authors following this scheme, and nearly all participants who
provided a codeable response under this coding scheme were coded as
approach. The authors and another researcher met to devise a new
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 513
method to coding these responses and decided to rate the amount of self-
efficacy participants had when considering how they would approach
new tasks. We used the following item to measure these responses: “To
what extent did the participant show an increase or decrease in self-
efficacy?” (1 = Participant described a strong decrease in self-efficacy, to
6 = Participant described a strong increase in self-efficacy, or N/A if the
participant’s response was not codeable under this scheme). An example
of a participant response for this third question was “This was the last
important task I did because it was extremely stressful and exhausting.
Retired immediately after.” Both authors coded this as a 1 and another
researcher coded this as a 2. Another response was “I felt that if I
achieved something so great that I could do just about anything I put my
mind to.”This was coded as a 5 and a 6 by the two authors respectively,
and another researcher coded this as a 6. The interrater reliability among
the two authors and another researcher on this measure was excellent, α =
.93. If at least two of the three coders were able to produce a rating on
this measure, the rating was averaged. There were 105 codeable
responses using this measure. On average, participants had a moderately
high degree of self-efficacy in their responses (M = 4.27, SD = 1.15).
We hypothesized these self-efficacy ratings would strongly correlate
with the quantitative ratings of self-efficacy and self-esteem and would
not correlate with NFD, similar to the Study 1 results. Interestingly, the
self-efficacy coded in these responses did not correlate significantly with
the quantitative measures of self-esteem (r = .06, p = .51) nor self-
efficacy (r = .12, p = .20). There was a weak marginal association
between NFD and these self-efficacy ratings, r = -.17, p = .09. This effect
size was similar to the relationship between NFD and the quantitative
measure of self-efficacy in this study. The qualitative self-efficacy
ratings did not correlate significantly with the coded positive emotions
nor negative emotions, r = .14, p =.15, and r = .10, p = .31, respectively.
Similar to the other exploratory analyses, we also assessed whether there
was a non-linear relationship between NFD and coder-rated self-efficacy.
Neither a quadratic model nor a cubic model significantly predicted
coder-rated self-efficacy, F(2, 102) = 1.74, p = .18 and F(3, 101) = 1.25,
p = .30, respectively.
Two Likert-type items followed the qualitative items – to what extent
participants believed they completed the task successfully (M = 5.20, SD
= 1.00; measured on a six-point scale), and how stressed they were taking
on the task (M = 3.75, SD = 1.28; measured on a five-point scale). These
were added as exploratory variables and we had no a priori hypothesis
about them. There was a weak marginal relationship between NFD and
the successful completion variable, r = -.16, p = .09. The qualitative self-
efficacy codings correlated moderately-to-strongly with the successful
514 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
completion variable, r = .40, p < .001 and interestingly, the stress
variable, r = .30, p = .002 (i.e. the more self-efficacy reported, the more
stress participants reported associated with the task). Also, the more
positive emotions reported, the greater participants reported having
completed the task successfully, r = .21, p = .03, and they reported less
stress, r = -.26, p = .006. Not surprisingly, the more negative emotions
coded, the more stress reported, r = .45, p < .001.
We also conducted exploratory analyses to see if there was a
curvilinear association between NFD and perceived successful task
completion, and between NFD and stress. There was a moderate U-
shaped effect of NFD predicting successful task completion, F(2,110) =
3.11, p = .05, R
2
= .05, similar to the effects of NFD predicting the
quantitative self-efficacy measure. A cubic model did not explain
additional variance predicting successful completions, F(1, 109) = 0.22,
p = .64. There was also a strong inverted-U-shaped effect of NFD
predicting stress, F(2, 110) = 7.27, p = .001, R
2
= .12 – As NFD
increases, recalled stress increased but decreased at the high end of NFD
(Figure 4). A cubic model did not explain additional variance in the
model predicting stress, F(1, 109) = 0.03, p = .98.
Figure 4 Quadratic Relationship between NFD and Recalled Stress when
Completing an Important Task
Discussion
We hypothesized that the relationship between self-esteem and self-
efficacy would change depending on an individual’s NFD personality –
that self-esteem and self-efficacy would not correlate as strongly at the
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 515
high end of NFD compared to lower levels. Despite the very strong
positive relationship between self-esteem and self-efficacy, and the
strong negative relationship between NFD and self-esteem, all the ways
that NFD relates to self-esteem has nothing to do with how self-esteem
relates to self-efficacy. Notably, in this study there was a weak
relationship between NFD and self-efficacy that was not found in Study
1. Perhaps including the self-esteem measure in the same study as self-
efficacy created a response bias, as the constructs are similar, and have
been argued to measure a common core construct of personality along
with neuroticism and locus of control (Judge et al., 2002). As individuals
responded to the self-esteem items, similar questions that measured self-
efficacy may have been responded to at the same strength since the
individuals were primed with similar concepts. To address this possible
confound, any future research in which these constructs appear together
should have a filler task in between the measures, or one of the measures
should be completed at a later time.
Judge et al. (2002) found that the four measures of self-esteem, self-
efficacy, neuroticism, and locus of control show poor discriminant
validity with one another. The zero or weak correlations found between
NFD and self-efficacy in this study and Study 1, and the moderately
strong correlation between NFD and self-esteem found in this study and
previous research (Frankowski et al., 2016) indicate that self-efficacy and
self-esteem can be discriminated by their relationship to NFD.
Notably, we replicated the curvilinear relationship between NFD and
self-efficacy that we found in Study 1’s M-Turk participants. Self-
efficacy decreases as NFD increases but self-efficacy increases at the
high end of NFD. This curvilinear relationship between NFD and self-
efficacy, however, does not explain why those at the high end of NFD
still tend have lower levels of self-esteem. Interestingly, in this study we
also found an inverted U-shaped relationship between NFD and recalled
stress when completing an important task. It seems a high NFD
personality may be a buffer to stress that can enhance self-efficacy, but
not self-esteem.
In the qualitative analysis, the moderate positive correlation between
perceived stress and self-reported self-efficacy appeared to run contrary
to what is expected. An individual typically interprets stress as signs of
vulnerability to poor performance (Bandura, 1994). This suggests one
should report lower self-efficacy with higher stress. To explain this
contradiction, we note that Bandura (1994) commented that it is
important to consider the person’s level of self-efficacy; those who have
high self-efficacy are likely to view their affective arousal as a facilitator
of performance, whereas those who have self-doubt consider that same
affective state as debilitating. Perhaps our Study 2 participants were
516 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
confident in their abilities to perform tasks or affect an outcome in their
favor, while perceiving their stress as a motivator. However,
interpretations about the correlation should be cautiously considered, as
people’s memories may be tainted by recall bias (Spencer et al., 2017).
GENERAL DISCUSSION
Across studies, NFD did not correlate with self-efficacy, but we
found support in two samples that the relationship is quadratic. This
relationship suggests that those at the high end of NFD may perceive
their manipulative and impulsive tendencies as self-efficacy. Persons
with positive self-concept, which includes self-efficacy, have been found
to be more motivated to perform their jobs (Judge et al., 1998). High
NFD individuals who consider their manipulative behaviors as self-
efficacy may honestly believe they can get things done. This sense of
accomplishment serves to sustain and socially motivate. When
manipulating other people gets these individuals what they desire, the
high NFD person may perceive this as self-efficacy. These perceptions of
self-efficacy possibly serve as a means of encouragement to remain
actively engaged in life activities, while also perpetuating the cycle of
drama to get what they want.
Future Directions
High NFD individuals do not appear to respond in a socially desirable
manner, based on the results of Study 1. It could be the case, though, that
high NFD individuals are still concerned with self-presentation in some
aspects of life that relate to self-efficacy. Research has shown that self-
efficacy and social desirability positively correlate among certain
populations, including teachers (Ashton et al.,1984) and college students
(Silverthorn & Gekoski, 1995). An upside to the NFD trait may be that
high NFD individuals in many careers, or those who are college students,
may be less apt to experience feelings of imposter syndrome (see
Kolligian & Sternberg, 1991) if they interpret challenges in their careers
and studies as victimizing, and in turn interpret their impulsive and
manipulative responses to such challenges as self-efficacy. The
curvilinear relationships between NFD and self-efficacy and NFD and
stress support this buffering effect of this personality type. This possible
upside of the NFD cycle should be investigated further.
Another future direction to explore is to experimentally manipulate
self-efficacy to determine whether NFD moderates aggression, internal
attributions of others, and dishonest behavior when individuals
experience a low versus high state of self-efficacy. If high NFD
individuals interpret their behaviors as self-efficacy, we may see an
overcompensation of dramatic tendencies when presented with a
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 517
situation that is meant to limit efficacy, e.g. an impossible task. There
may be downstream effects in which high NFD behavior intensifies in
response to a threat to self-efficacy.
Conclusion
In two studies with three samples, we found there is no linear
association between NFD and self-efficacy, but that there is a curvilinear
relationship such that those at the low and high ends of NFD have higher
levels of self-efficacy compared to those in the mid-range of the measure.
Need for Drama does not appear to moderate the relationship between
self-esteem and self-efficacy – those at the high end of the NFD measure
tend to have higher levels of self-efficacy compared to those in the
middle of the NFD measure, but they tend to also have lower levels of
self-esteem. Their cognitive appraisals of themselves do not align with
their perceived abilities to accomplish tasks. We believe a plausible
explanation for our findings is that those high in NFD are perceiving
their impulsive and manipulative behaviors as self-efficacy, as it serves
to sustain, motivate, and buffer them from stress.
Acknowledgments: The authors wish to thank Teresa C. Tempelmeyer, from the
Department of Psychology at Midwestern State University, for assisting in the
qualitative analysis.
Declaration of Conflicting Interests: The authors declare that there are no
conflicts of interest.
Funding: This research did not receive any specific grant from funding agencies
in the public, commercial, or not-for-profit sectors.
REFERENCES
Arnett, J. J. (2006). Emerging adulthood: Understanding the new way of coming
of age. In J. J. Arnett & J. L. Tanner (Eds.), Emerging adults in America:
Coming of age in the 21st century (pp. 3-19). American Psychological
Association. https://doi.org/10.1037/11381-001
Ashton, P., Buhr, D., & Crocker, L. (1984). Teachers’ sense of efficacy: A self-
or norm-referenced construct? Florida Journal of Educational Research,
26(1), 29-41.
Bandura, A. (1994). Self-efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of
human behavior (Vol. 4, pp. 71-81). New York, NY: Academic Press.
(Reprinted in H. Friedman [Ed.], Encyclopedia of mental health. San Diego,
CA: Academic Press, 1998).
Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P. Robinson,
P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social
psychological attitudes: Volume 1 of measures of social psychological
attitudes series (pp. 115-160). San Diego, CA: Academic Press, Inc.
Brehm, J. W. (1999). The intensity of emotion. Personality and Social
Psychology Review, 3(1), 2-22. https://doi.org/10.1207
/s15327957pspr0301_1
518 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
Burdenski, T. K., Jr. (2000, April 24-28). Evaluating univariate, bivariate, and
multivariate normality using graphical procedures [Paper presentation].
American Educational Research Association Annual Meeting, New Orleans,
LA, United States. https://eric.ed.gov/?id=ED440989
Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics
to nonnormality and specification error in confirmatory factor analysis.
Psychological Methods, 1(1), 16-29. https://doi.org/10.1037/1082-
989X.1.1.16
Ewen, R. B. (2014). An introduction to theories of personality, Seventh edition,
New York, NY: Psychology Press.
Fahey, G. (2018). Faking good and personality assessments of job applicants: A
review of the literature. DBS Business Review, 2, 45-68.
https://doi.org/10.22375/dbr.v2i0.25
Frankowski, S., Lupo, A. K., Smith, B. A., Dane'El, M., Ramos, C., & Morera,
O. F. (2016). Developing and testing a scale to measure need for drama.
Personality and Individual Differences, 89, 192-201.
https://doi.org/10.1016/j.paid.2015.10.009
Furr, R. M. (2010). Social desirability. In N. J. Salkind (Ed.), Encyclopedia of
research design (pp. 1396-1398). Thousand Oaks, CA: SAGE Publications,
Inc. https://doi.org/10.4135/9781412961288.n425
Hartman, R. O., & Betz, N. E. (2007). The five-factor model and career self-
efficacy: General and domain-specific relationships. Journal of Career
Assessment, 15(2), 145–161. https://doi.org/10.1177/1069072706298011
Hayes, A. F. (2018). Introduction to mediation, moderation, and conditional
process analysis: A regression-based approach, Second edition, New York,
NY: The Guilford Press.
Heydarian, N. M. (2016). Developing theory with the grounded-theory approach
and thematic analysis, Observer, 29(4).
https://www.psychologicalscience.org/observer/developing-theory-with-the-
grounded-theory-approach-and-thematic-analysis
Judge, T. A., & Bono, J. E. (2001). Relationship of core self-evaluations traits—
self-esteem, generalized self-efficacy, locus of control, and emotional
stability—with job satisfaction and job performance: A meta-analysis.
Journal of Applied Psychology, 86(1), 80-92. https://doi.org/10.1037//0021-
9010.86.1.80
Judge, T. A., Erez, A., & Bono, J. E. (1998). The power of being positive: The
relation between positive self-concept and job performance. Human
Performance, 11(2-3), 167-187. https://doi.org/10.1080/08959285.
1998.9668030
Judge, T. A., Erez, A., Bono, J. E., & Thoresen, C. J. (2002). Are measures of
self-esteem, neuroticism, locus of control, and generalized self-efficacy
indicators of a common core construct? Journal of Personality and Social
Psychology, 83(3), 693-710. https://doi.org/10.1037/0022-3514.83.3.693
Kolligian, J., Jr., & Sternberg, R. J. (1991). Perceived fraudulence in young
adults: Is there an 'imposter syndrome'? Journal of Personality Assessment,
56(2), 308-326. https://doi.org/10.1207/s15327752jpa5602_10
Kowalski, C. M., Rogoza, R., Vernon, P. A., & Schermer, J. A. (2018). The dark
triad and the self-presentation variables of socially desirable responding and
Lue & Frankowski DRAMA PERCEIVED AS SELF-EFFICACY 519
self-monitoring. Personality and Individual Differences, 120, 234-237.
https://doi.org/10.1016/j.paid.2017.09.007
McCrae, R. R., & Costa, P. T. (1986). Personality, coping, and coping
effectiveness in an adult sample. Journal of Personality, 54(2), 385-404.
https://doi.org/10.1111/j.1467-6494.1986.tb00401.x
Pantaleo, G., Miron, A. M., Ferguson, M. A., & Frankowski, S. D. (2014).
Effects of deterrence on intensity of group identification and efforts to protect
group identity. Motivation and Emotion, 38, 855-865.
https://doi.org/10.1007/s11031-014-9440-3
Paulhus, D. L., & Williams, K. M. (2002). The dark triad of personality:
Narcissism, Machiavellianism, and psychopathy. Journal of Research in
Personality, 36(6), 556-563. https://doi.org/10.1016/S0092-6566(02)00505-6
Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for
probing interaction effects in multiple linear regression, multilevel modeling,
and latent curve analysis. Journal of Educational and Behavioral Statistics,
31(4), 437-448. https://doi.org/10.3102/10769986031004437
Reynolds, W. M. (1982). Development of reliable and valid short forms of the
Marlowe-Crowne social desirability scale. Journal of Clinical Psychology,
38(1), 119-125. https://doi.org/10.1002/1097-4679(198201)38:1<119::AID-
JCLP2270380118>3.0.CO;2-I
Robinson, J., Rosenzweig, C., Moss, A. J., & Litman, L. (2019). Tapped out or
barely tapped? Recommendations for how to harness the vast and largely
unused potential of the Mechanical Turk participant pool. PLoS ONE,
14(12),1-29. https://doi.org/10.1371/journal.pone.0226394
Rosenberg, M. (1965). Society and the adolescent self-image, Princeton, NJ:
Princeton University Press.
Schwarzer, R., & Jerusalem, M. (1995). Generalized self-efficacy scale. In J.
Weinman, S. Wright, & M. Johnston (Eds.), Measures in health psychology:
A user’s portfolio. Causal and control beliefs (pp. 35-37). Windsor, England:
NFER-NELSON.
Spencer, E. A., Brassey, J., & Mahtani, K. (2017). Recall bias. Catalogue of
Bias. https://catalogofbias.org/biases/recall-bias/
Silverthorn, N. A., & Gekoski, W. L. (1995). Social desirability effects on
measures of adjustment to university, independence from parents, and
self‐efficacy. Journal of Clinical Psychology, 51(2), 244-251.
https://doi.org/10.1002/1097-4679(199503)51:2<244::AID-JCLP2270510214
>3.0.CO;2-Q
James C. Lue https://orcid.org/0000-0003-4114-6834
Scott D. Frankowski https://orcid.org/0000-0003-1820-2615
520 NORTH AMERICAN JOURNAL OF PSYCHOLOGY
