ArticlePDF Available

Seeing the invisible: Extracting signs of depression and suicidal ideation from college students’ writing using LIWC a computerized text analysis

Authors:

Abstract and Figures

Depression remains one of the leading problems around the world. Partly, the problem stems from the fact that depression is still difficult to diagnose using traditional assessment tools. A few pieces of evidence suggest that text analysis is capable of identifying psychological states and mental issues including depression. However, this method has not been tested in the Philippines. The main purpose of this study, therefore, was to determine clues of depression and suicidal ideation in college students’ writing using the online version of LIWC, a computerized text analysis. The present study was conducted at Central Luzon State University, Philippines where 159 undergraduate students participated. Using correlational analysis, several important findings were found. First, college students with a high level of depression and suicidal ideation tend to write more personal pronouns such as “I”, “me”, and “my” in their writing. Second, LIWC traditional dimensions such as I-words and negative emotions show a significant link with depression but only the I-words dimension was significantly associated with suicidal ideation. Third, LIWC summary variables such as clout and emotional tone also show a significant relationship with depression and suicidal ideation. Furthermore, regression analysis suggests that these LIWC domains were significant predictors of depression. However, only emotional tone was found to have a modest but significant influence on suicidal ideation. Lastly, the length of the essay was significantly correlated with depression but not with suicidal ideation. The theoretical and practical implications of the findings are discussed.
Content may be subject to copyright.
International Journal of Research Studies in Education
2020 Volume 9 Number 4, 31-44
© The Author(s) / Attribution CC BY
Seeing the invisible: Extracting signs of depression and
suicidal ideation from college students’ writing using
LIWC a computerized text analysis
Lumontod, Robinson Z., III
Central Luzon State University, Philippines (robinsonlumontodiii@gmail.com)
Received: 30 January 2020 Revised: 4 April 2020 Accepted: 2 May 2020
Available Online: 29 June 2020 DOI: 10.5861/ijrse.2020.5007
ISSN: 2243-7703
Online ISSN: 2243-7711
OPEN ACCESS
Abstract
Depression remains one of the leading problems around the world. Partly, the problem stems
from the fact that depression is still difficult to diagnose using traditional assessment tools. A
few pieces of evidence suggest that text analysis is capable of identifying psychological states
and mental issues including depression. However, this method has not been tested in the
Philippines. The main purpose of this study, therefore, was to determine clues of depression
and suicidal ideation in college students’ writing using the online version of LIWC, a
computerized text analysis. The present study was conducted at Central Luzon State
University, Philippines where 159 undergraduate students participated. Using correlational
analysis, several important findings were found. First, college students with a high level of
depression and suicidal ideation tend to write more personal pronouns such as “I”, “me”, and
“my” in their writing. Second, LIWC traditional dimensions such as I-words and negative
emotions show a significant link with depression but only the I-words dimension was
significantly associated with suicidal ideation. Third, LIWC summary variables such as clout
and emotional tone also show a significant relationship with depression and suicidal ideation.
Furthermore, regression analysis suggests that these LIWC domains were significant
predictors of depression. However, only emotional tone was found to have a modest but
significant influence on suicidal ideation. Lastly, the length of the essay was significantly
correlated with depression but not with suicidal ideation. The theoretical and practical
implications of the findings are discussed.
Keywords: depression; suicidal ideation; LIWC; text analysis; college students
Lumontod, R. Z., III
32 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
Seeing the invisible: Extracting signs of depression and suicidal ideation from college
students’ writing using LIWC a computerized text analysis
1. Introduction
Depression is a worldwide epidemic that affects the large percentage of the population. The prevalence of
depression is far more increasing in student population around the world. In fact, in Africa alone, almost half of
the student population suffer from depression (Tam, Lo, & Pachero, 2018). The number can be even more
alarming around the world. Unfortunately, only a small percentage received the appropriate intervention. The
sheer amount of help received by the affected ones might stem from two main issues; (a) because depression is,
in most cases, invisible. In effect, any effort to examine the number of people affected by depression is often
inaccurate. As a result, the available data on depression is only the tip of the iceberg. More students are affected
than the reported number (Rotenstein, et al., 2016). This leads educators and mental health professionals unable
to respond to the problem effectively. Even school records could not accurately assess depression (Kuo, Stoep,
Herting, Grupp, & McCauley, 2013). As a result, many depressed students remain untreated. When this
happened, suicide cases may become even more prevalent. In the Philippines for instance, suicide cases in
adolescents and young adults are increasing (Quintos, 2017). On the other study, Quintos (2017) finding
suggested that approximately one in every ten Filipino youth had suicidal thoughts. The reported rate of suicidal
ideation in the Philippines is higher than the neighboring Asian counties such as Malaysia where only 6.3% were
reported to have suicide ideation (Maniam et al., 2013). Previous studies had provided clues on the current
psychological states of youths around the world. The need for urgent action is clear. The challenge for teachers,
academicians, and school administrators is to correctly identify depressed students and give corresponding and
appropriate intervention. However, the process can be extremely difficult. Partly because depression is often
concealed and repressed by the bearer. Generally, because of the stigma of being labeled as “depressed” or
“psychologically ill” is a significant reason why most depressed students do not truly share their psychological
turmoil.
Furthermore, the other hindrance in diagnosing depression is the assessment tools being used. To date, most
depression assessments rely on traditional techniques. This paper argues that, while depression assessment tools
are helpful, their ability to yield accurate results is weak. Social desirability and the willingness of the
participants to report their inner experience often negatively affects the participants’ quantitative responses.
Other techniques should be developed to complement the traditional tools. Fortunately, modern techniques are
available today. One of them is text analysis. This computer-aided tool yielded promising results that could
potentially guide the more accurate assessment of the psychological states. Text analysis does not rely on biased
self-reports but on the themes that are prevalent in personal writings. Studies on this area, however, still dearth,
especially in the Philippine setting. If personal writings could provide clues of depression and suicidal ideation,
then using it as an assessment tool may lessen the increasing incidence of suicides in Filipino youths.
Text analysis, as it is commonly known, is not new. Previous studies had found how written language
reflects individuals’ mental state (Boyd & Pennebaker, 2015); social status (Kacewics, Pennebaker, Davis, Jeon,
& Graesser, 2013); lies (Newman, Pennebaker, Berry, & Richards, 2003). However, most previous studies in this
field had mainly focused on the political, sociological, and pure linguistic arena. Therefore, the purpose of this
study was to identify clues of depression and suicidal ideation in personal writing using text analysis.
1.1 Significance of the study
This paper might be the first attempt to use personal writing analysis to identify cues of depression and
suicide ideation in the Philippine setting. The result of this study will greatly help not only in identifying signs of
depression and suicidal ideation but also in designing new strategies in assessing psychological issues. Mental
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 33
health professionals and educators will have an alternative way of identifying students with depression and
suicidal ideation.
2. Review of related literature
2.1 Language and psychological states
For decades, researchers had shifted their focus of investigation on the importance of language in
understanding psychological states. A handful of empirical efforts produced some significant contributions to
understanding individuals’ behavior. Consistently, the findings of the previous body of literature suggest that,
indeed, words reflect the inner world of an individual. For instance, Boroditsky (2001) suggests that language
could shape the individuals’ thinking process. Personal stories and psychological states seem to relate to one
another. Vaughn (2018) found similar interesting finding which suggests that when people think about hope, their
stories reflect positive experiences. This may further imply that the spoken words hide more meanings than what
they literally mean. In fact, words could reveal lies (Newman, Pennebaker, Berry, & Richards, 2003), political
ideologies (Chichoka, Bilewics, Jost, Marrouch, & Witkowska, 2016), and political trends (Jordan, Sterling,
Pennebaker, & Boyd, 2019). Beyond politics, language also reflects students’ performance in school. Pennebaker,
Chung, Frazee, Lavergne, and Beaver (2014) found that the seemingly irrelevant small words may predict
students’ academic performance. Armed with these contentions, this paper assumes that words may also provide
clues of psychological states in the Philippine context.
The previous finding showed a promising method in understanding individuals’ cognitive and behavioral
aspects. Understanding the hidden meanings behind the spoken and written words provide a window to a new
way of discovering and/or predicting human behavior. The challenge commonly found in diagnosing depression
can be remedied.
2.2 Depression
The prevalence of depression around the world is increasingly alarming. Unfortunately, depression not only
affects the general population but more so in students. A recent study shows that almost half of the student
population in Africa alone suffer from depression (Tam, Lo, & Pachero, 2018). This might be the result of
academic demands that college students constantly dealing on a daily basis. As college students progress in their
academic year level, they tend to become more depressed (Horgan, Sweeney, Behan, & McCarthy, 2016). The
academic requirements for each year level may have contributed to the increase in depression level. On the other
hand, Chen et al. (2015) found that students’ depressive symptoms were linked to academic performance. As
students become more depressed, their academic performance suffers. The effect of depression can be
devastating in students’ overall functioning (Nyer, et al., 2013). As per available research findings, medical
students were consistently diagnosed with high depression level. In fact, they are more depressed than the
general population (Dahlin, Joneborg, & Runeson, 2005). The findings of Puthran, Zhang, Tam, and Ho (2016)
also suggest that about 28% of medical students around the world suffered from depression. The disproportion
between sexes was also noted. Female students have higher depression than males.
The determining factors were identified as to why medical students get depressed. The factors include poor
academic status and dissatisfaction with the medical education received (Ediz, Ozcakir, & Bilgel, 2017). Being
hopeless of the situation might also cause depression among medical students (Coskun, Ocalan, Ocbe, Semiz, &
Budakoglu, 2019). The body of literature on depression suggests one thing; students’ mental health is at stake.
Students’ mental health is dwindling and needs to be addressed appropriately (Castillo & Schwartz, 2013).
However, regardless of the population affected, appropriate assessment and intervention techniques certainly
needed. As reported in many studies, depression, if not treated properly may result in several psychological
issues, especially in college students. One of which is suicidal ideation.
Lumontod, R. Z., III
34 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
2.3 Suicidal ideation
Suicidal ideation is one of the byproducts of depression. In fact, depression symptoms were found to have a
significant link with suicide ideation among college students (Garlow et al., 2008). Previous studies had
identified risk factors of suicide ideation. One study found that depressive symptoms and feelings of
hopelessness both predict suicide ideation (Lamis, Ballard, May, & Dvorak, 2016). However, social support is a
moderating factor that moderates the relationship between depressive symptoms and hopelessness in influencing
suicidal ideation. Previous studies, on the other hand, found important findings. Dissatisfaction in daily living
seems to be a predictor of suicidal thoughts. Tan, Xia and Reece (2016) argued that the standard of living was
associated with suicidal ideation. Much like depression, suicide ideation is shaped by contextual dynamics.
However, social context is a macro factor that oftentimes more complicated to apply in making sense as to why
people have developed suicidal thoughts. Fortunately, previous studies had explored different factors that are
also predictors of suicide ideation. Most of the identified factors are behavioral and psychological. For instance,
Dunn, Goodrow, Givens, and Austis (2008) suggest that substance abuse and suicidal ideation did have a
significant link. Delfabbro, Mavaso, Winefield, and Winefield (2015) found that smoking, romantic relationship
and poor health predict suicide ideation.
Aside from the behavioral factors, the thinking process also found to be a risk factor of suicide ideation. The
previous study of Teismann and Forkmann (2015) reflects that the mere perception of entrapment may predict
suicide ideation. Although depression is a strong predictor of suicide ideation, certainly, there are other factors
that also suicide ideation risk. At least, one study suggests that poor family cohesiveness could possibly be a
suicide ideation risk factor (Wong, Uhm, & Li, 2012). Negative perceptions about oneself can be a relevant
factor in suicidal thoughts. Psychological turmoil is one if not the strongest predictor of suicide ideation and
suicide attempts. Shneidman’s model of suicide ideation seems to provide a clue on what triggers suicidal
thoughts. The previous study found out that psychache or psychological pain (Shneidman, 1993) was found to
have had a significant link with suicide ideation (Troister, Davis, Lowndes, & Holden, 2013). Recent findings
seem to strengthen the contention on the relation between psychache and suicide ideation. Montemarano,
Troister, Lambert, and Holden (2018) argued that all factors are relevant to suicide ideation only if they linked
through psychache. If this contention can be held true, then psychological pain is the strongest determining
factor of suicide ideation. However, although previous findings shed light on the nature and relevant factors of
depression, one problem remains. To date, identifying symptoms of depression and suicidal ideation remains a
grand challenge for mental health professionals. This study argues that knowing the driving factors of
psychological problems is one thing; identifying them is another.
2.4 Current study
As mentioned, influential factors of depression and suicide ideation are almost clear. The remaining problem,
as most mental health professionals had found in the process of identifying signs of depression and suicide
ideation. The main purpose of this study, therefore, is to identify the sign of depression and suicide ideation
embedded in personal writing using text analysis. Because this technique focuses on words that are directly or
indirectly linked to depression and suicidal ideation, social desirability and other response biases were
minimized. This paper further argued that personal writing could provide clues of depression and suicidal
ideation. Hence, the primary focus of this inquiry was to analyze the personal writings of college students. This
study may be the first to use text analysis specifically LIWC in the Philippine setting in identifying words
associated with depression and suicide ideation. The result of this endeavor will be of great help to mental health
interventions by providing keys to identifying signs of psychological problems.
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 35
3. Method
3.1 Design
This study used correlational research design in determining the relationship and predictive ability of LIWC
domains on depression and suicidal ideation. Each variable was measured using the following assessment tools.
Linguistic Inquiry and Word Count (LIWC) - The online version of LIWC was used to analyze the
personal writings of the participants. It is a software that calculates the number of words, and grouped them in
several psychological categories. This program identifies the number of important words from personal writings
and groups them into two main categories. The first one is the Traditional LIWC dimension which reflects the
percentage of words present in each dimension. Under this category, there are five dimensions which include
I-Words (I, Me, My), Social Words, Positive Emotions, Negative Emotions, and Cognitive Processes. Because the
output was in percentage, the number of words in each dimension was identified by multiplying the percentage
from the total number of words that each participant was written.
The second major category is the Summary Variables. This category has four domains which include
Analytic, Clout, Authenticity, and Emotional Tone. The analytic domain pertains to formal thinking; clout
pertains to authoritativeness, confidence, and leadership capabilities; authenticity refers to the truthfulness of the
writing; emotional tone refers to the emotional state of the participant based on the rated text. Unlike the
traditional dimension, each domain of the summary variables uses a 100-point scale (0 = Low, 100 = High) to
determine the level of each domain.
Beck’s Depression Inventory (BDI) The freshmen students’ academic adjustment was measured using
Beck’s Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961). It is a 21-item questionnaire
that has a 4-point scale. The total score is the sum of all scores of all items which determine the level of
depression (1-10 = The ups and downs are considered normal; 11-16 = Mild mood disturbance; 17-20 =
Borderline clinical depression; 21-30 = Moderate depression; 31-40 = Severe depression; over 40 = Extreme
depression). Beck, Steer and Carbin (1988) reported an acceptable internal consistency for the scale (α = .81) for
non-psychiatric participants. In the present study, BDI shows higher internal consistency (α = .87).
Suicidal Ideation Attributes Scale (SIDAS) - To measure the participants’ suicide ideation, the Suicidal
Ideation Attributes Scale (Van Spijker et al., 2014) was used. This scale has five items that assess the suicide
ideation of the participants. Each item was scored using a 10-point scale. It has a reported internal consistency of
(α = .91). In this study, SIDAS also yielded higher internal consistency (α = .94).
3.2 Participants
There were 176 second-year psychology students of Central Luzon State University participated in this
study. However, the initial examination of the data revealed some missing values and inconsistencies that could
negatively affect the outcome of the analysis. To obtain cleaner data, 17 participants who had incomplete
responses (and other irregularities) were discarded from the data set. Only 159 participants were included in this
study with age ranging from 19-23 years old (M = 18.87, SD = .83).
Prior to the data gathering, the proposed topic underwent a review by the Institutional Review Board (IRB)
in the university to make sure that all ethical standards and considerations were taken into consideration. Upon
receiving the permission to proceed, the data gathering immediately commenced. The data gathering had two
parts. First, the participants were instructed to respond to Beck’s Depression Inventory and the Suicide Ideation
Attributes Scale. The first step was conducted in a classroom environment were students could comfortably
respond to the questionnaires. After completed the questionnaires, the participants were then instructed to do the
second part which was essay writing. They were told to write an essay about their current emotional feelings in
Lumontod, R. Z., III
36 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
the English language. The essay has a minimum 100 words and a maximum of 500 hundred words (this is due to
the limitation of the online version of LIWC). To give the participants enough time to contemplate, the
researcher allowed them to bring the task home. The finished essays were submitted via email several days later.
The participants who completed the tasks were given extra credits.
3.3 Ethical considerations
No participant in this study was forced to respond to the questionnaires and write the essay. In the same
manner, this study did not cause any emotional or psychological uneasiness to the participants. All participants
were given the freedom to withdraw from participating in the study before, during, and after the data gathering
process. Anonymity and confidentiality were both assured.
3.4 Data analyses
The Pearson Moment Correlation was used in analyzing the relationship between the Traditional LIWC
Dimensions, Summary Variables, depression level, suicidal ideation, and the total number of words. Stepwise
regression analyses were conducted to determine the best predictor domains of Traditional LIWC Dimensions
and Summary Variables on depression and suicide ideation.
4. Results
This study aimed to identify signs of depression and suicidal ideation from personal writings. Using the
Linguistic Inquiry and Word Count (LIWC) results in traditional dimensions and summary variables were
correlated with depression and suicidal ideation. Important findings are indicated in the following tables.
Table 1
Relationship between traditional LIWC dimensions, summary variables, words of the essay and BDI with SIDAS
Items BDI SIDAS
r p r p
Traditional LIWC Dimensions
I-Words (I, Me, My) .289** .000 .202* .011
Social Words .028 .727 .023 .771
Positive Emotions -.105 .186 -.121 .13
Negative Emotions .280** .000 .142 .075
Cognitive Processes .131 .100 .099 .212
Summary Variables
Analytic -.078 .328 -.010 .899
Clout -.213** .007 -.166* .037
Authenticity .093 .242 .041 .611
Emotional Tone -.295** .000 -.213** .007
Total Number of Words of the Essay .166* .036 .147 .065
Note. *p < .05, **p < .01, ***p < .001.
The correlational analysis in table 1 revealed some modest but significant findings. First, out of five
traditional LIWC dimensions, only the number of I-words used (r = .289, p = .000) and negative emotions (r
= .280, p = .000) shown significant relationship with depression. With SIDAS, however, only I-words was
significantly correlated (r = .202, p = .011).
Second, of four summary variables, only two had consistent significant correlation with depression and
suicide ideation. Clout was significantly linked with depression (r = -.213, p = .007) and suicide ideation (r =
-.166, p = .037). Likewise, emotional tone also shown significant relationship with depression (r = -.295, p
= .000) and suicidal ideation (r = -.213, p = .007). Lastly, the total number of words of the essay was
significantly correlated with depression (r = .166, p = .036).
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 37
Table 2
Regression analysis on the influence of traditional LIWC dimensions on depression and suicidal ideation
Traditional LIWC Dimension BDI
SIDAS
B SE B β
B SE B β
Step 1
Constant 15.721 1.648 5.385 2.262
I-Words .151 .040 .29** .142 .055 .202*
R
2
.084 .041
F 14.346**
6.653*
Step 2
Constant 17.661 1.603 7.646 2.242
I-Words .277 .046 .53** .288 .065 .410**
Positive Emotions -.572 .122 -.413**
-.666 .171 -.359**
R
2
.197 .126
F 19.110**
11.249**
R
2
.113 .085
F 21.959 15.241
Step 3
Constant 16.859 1.639
I-Words .223 .053 .425**
Positive Emotions -.579 .121 -.418**
Negative Emotions .246 .125 .177*
R
2
.217
F 14.279**
R
2
.020
F 3.905
Note. *p < .05, **p < .01, ***p < .001, B = unstandardized regression coefficient; SE = standard error; ß = standardized regression
coefficient; = difference in the proportion of the explained variance.
In table 2, the predictive ability of the Traditional LIWC Dimensions on depression and suicide ideation was
tested using stepwise regression. The result shows important unique findings. Out of five traditional LIWC
dimensions, only three were significant predictors (I-words, positive emotions, and negative emotions) and when
combined yielded a significant influence on depression (F (3, 155) = 14.279, p = .000).
As indicated in the table, the I-words domain was found to be a significant predictor on depression (R
2
= .084,
F (1, 157) = 14.346, p = .000). About 8.4% of the depression variance can be attributed to I-words which is a
significant predictor of depression (β = .29). The Positive emotions dimension was found to be the second
significant predictor (β = -.413, p = .000) which increased the variance of depression by 11.3% making it the
second significant predictor of depression (R
2
= .113, F (1, 156) = 21.959, p = .000). The last significant predictor
was negative emotions which potentially influenced the depression variance by 2.0%. Although modest, negative
emotions significantly influenced depression (R
2
= .020, F (1, 155) = 3.905, p = .050) and a potential predictor (β
= .177, p = .050).
Table 2 also shows the predictive ability of the Traditional LIWC dimensions on suicide ideation. Out of five
domains, only two were significant predictors (I-words and positive emotions domains) which had the significant
predictive ability (F (1, 156) = 11.249, p = .000).
The I-words domain was found to be a significant predictor of suicidal ideation (β = .202, p = .011) and
potentially influenced the variance of suicidal ideation by 4.1% (R
2
= .041, F (1, 157) = 6.653, p = .011). Positive
emotions domain was also found to have a predictive ability on suicidal ideation (β = -.359, p = .000) and
influenced the suicide ideation variance by 8.5% (R
2
= .085, F (1,156) = 15.241, p = .000).
Lumontod, R. Z., III
38 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
Table 3
Regression analysis on the influence of LIWC summary variables on depression and suicidal ideation
LIWC Summary Variables BDI
SIDAS
B SE B β
B SE B β
Step 1
Constant 24.312 1.093
13.519 1.499
Emotional Tone -.091 .024 -.295**
-.089 .032 -.213**
R
2
.087
.045
F 14.925**
7.449**
Step 2
Constant 25.720 1.211
Emotional Tone -.087 .023 -.279**
Clout -.100 .040 -.190*
R
2
.123
F 10.895**
R
2
.036
F 6.355
Note. *p < .05, **p < .01, ***p < .001, B = unstandardized regression coefficient; SE = standard error; ß = standardized regression
coefficient; = difference in the proportion of the explained variance.
The table 3 indicates the results of the stepwise regression analysis on the predictive ability of LIWC
summary variables on depression and suicide ideation. Out of four variables, only two (emotional tone and clout)
were identified as significant predictors of depression (F (2, 156) = 10.895, p = .000). Individually, emotional
tone might influence depression’s variance by 8.7% (R
2
= .087, F (1, 157) = 14.925, p = .000) making it a
significant predictor on depression (β = -.295, p = .000). Likewise, clout also contributed to the variance of
depression by 3.6% (R
2
= .036, F (1, 156) = 6.355, p = .013). However, out of four summary variables, only
emotional tone was a significant predictor of suicidal ideation (F (1, 157) = 7.449, p = .007). Although only 4.5%
of the variance of suicidal ideation can be attributed to emotional tone, it is a highly significant predictor (β = -.213,
p = .007).
5. Discussions
The main purpose of this study was to fill the gap in the literature in identifying signs of depression and
suicidal ideation in personal writings. This study found important findings that could help mental health
professionals to identify depression individuals using text analysis. The main findings and implications of the
results are discussed.
5.1 The relationship between LIWC dimensions and depression with suicide ideation
The result shows that out of five traditional LIWC dimensions (I-words, social words, positive emotions,
negative emotions, cognitive processes), only I-words and negative emotions dimensions were found to have a
modest but significant link with depression. The findings may suggest two interesting points. First, the use of
personal pronouns in personal writing may suggest psychological states inherent in the writer’s inner experience.
Although modest in strength, the correlation results provide a promising way of understanding depression.
Certainly, the findings need further verification. However, it might be plausible to assume that depression can be
seen by seeing the invisible cues embedded in the person’s written language. The present result corroborates
previous studies that show the significant association between depression and personal pronouns (Bernard,
Baddeley, Rodriguez, & Burke, 2015; Brockmeyer, et al., 2015). In addition, the finding of this study supports
the contention of Pennebaker, Mehl, and Niederhoffer (2003) which suggests that psychological states can be
identified through the individuals’ written words. However, the result did not account for the possible
methodological flaws of the study.
One of the weaknesses of the finding stems from the data gathering procedure wherein participants were
instructed to write an essay about their current personal emotional feeling. The essays, would, of course, pertain
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 39
to the personal account and therefore would more likely to produce personal pronoun words or I-words.
However, the current finding may somehow show that, even in a collectivist culture, the use of personal
pronouns is eminent in those with psychological turmoil. Do people in different cultural and social contexts
follow a similar pattern in writing when experience depression? This may be another question that deserves
future scientific inquiry. Furthermore, this finding adds support to the existing evidence which shows the
possible application of text analysis in identifying the hidden themes of psychological issues in personal writings
often missed by traditional assessment tools.
Secondly, the negative emotions dimension also shows a highly significant relationship with depression.
Although it is not surprising, this finding somehow strengthens the assumption that the LIWC detects
psychological states. In this case, the number of words related to negative emotions was found to have a
significant relationship with depression. Again, because this might be the first to use LIWC in analyzing personal
writings in the Philippine setting, future research endeavors may be needed to verify the current finding.
Nonetheless, this study somehow provides a new window for clinicians and mental health professionals in
determining signs of depression in a whole new way. If the findings of this study can be held ubiquitous,
therefore, the present results can be used in other social and cultural contexts.
On suicide ideation, however, only the number of personal pronouns (I-words) was significantly correlated.
The finding further suggests that personal pronouns may also a sign of suicide ideation. The more personal
pronouns found in personal writing, the more likely the writer experience suicidal thoughts. This might be the
first study that used LIWC in identifying suicidal ideation cues. Like depression, suicide ideation is often
concealed and often undetected by traditional assessment tools. Using text analysis especially LIWC may be of
great help in unraveling mental health issues. Personal pronouns are indicative of both depression and suicidal
ideation. This study does not only shed new light on clinical practice but also in formulating preventive actions.
Moreover, out of four LIWC summary variables, clout and emotional tone were the only domains
significantly correlated with depression. The finding may imply that depressed individuals may show low
authority related words in their writing. This is reflected in the negative correlation between clout and depression.
At least one study seemed to provide a vital clue. Pudrovska and Karraker (2014) contend that depressive
symptoms were high among women with job authority than those without job authority. The reason behind, as
the previous authors argued, was the cultural sense of femininity and masculinity. In this study, however, it is not
clear whether a similar explanation is possible. Context wise, previous studies focused heavily on job authority
and not the sense of authoritativeness depicted by the clout domain of LIWC. To this point, the result of the
present study could only show statistical significance but the explanation remains dearth. Future research on this
field may further clarify latent factors that could provide a meaningful explanation.
The same result found between emotional tone and suicidal ideation. Emotional tone refers to the emotional
state of the writer reflected in personal writing. As the result suggests, individuals with suicidal thoughts showed
lower scores on the emotional tone domain. This might support the existing empirical evidence which suggests
that expressive writing helps improve psychological health (Niles, Haltom, Mulvenna, Lieberman, & Stanton,
2014). As the college student wrote their essay about their current emotional state, their emotional turmoil might
have decreased over the course of the whole writing process. As depression level dwindles, suicidal thoughts
associated with depression might have also decreased.
Another interesting result of this study suggests that the length of personal writing was significantly
correlated with depression. Although it was low, a significant relationship may suggest that depressed individuals
tend to write longer when describing their emotional state. To the knowledge of the author, this is the first effort
to document the association between length of personal writing and depression. Apart from using computerized
tools to unravel depression cues in personal writings, looking at the actual length of the essay or any written
language may already provide clues. But how long personal writing becomes indicative of depression? This
question is beyond the scope of this study. The participants of this study were instructed to write a 100-500-word
Lumontod, R. Z., III
40 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
essay about their current emotional feelings. Setting a minimum and maximum amount of words was primarily
due to the limitation of the online LIWC version. Lower than 50 words, as the guideline of the online version,
indicates, may not yield an accurate result. Since the online version only reads the first 500 words of the text, the
maximum number of words was set. This paper, however, suspects that, while minimum and maximum word
requirement for the essay was important in the study, it too could detriment the accuracy of the statistical result.
Two things might have happened. First, participants were forced to write a hundred words longer than they
would willingly write. Second, participants who would have written longer essays were constrained and
therefore forced to limit the length of the essay. Both cases may have had affected the obtained data and the
statistical result as well. Future research on this topic may try to allow participants to write an essay freely. This
study also recommends the use of the paid version of LIWC.
5.2 The influence of LIWC traditional dimensions on depression and suicidal ideation
The influence of traditional dimensions on depression and suicidal ideation was assessed. The finding
suggests that personal pronouns (I-words), positive and negative emotions significantly predicted depression.
Consistent with the correlation result, depressed individuals tend to write more personal pronouns than those
with lower depression levels. This strengthens the assumption that I-words could be, indeed, the sign of
depression. This assumption, however, needs to be verified and owes future research undertaking. Nonetheless,
the present finding corroborates with the previous contention of Chung and Pennebaker (2007) which stated that
regular words used in daily communication reflect the individual’s psychological states. This finding may be
tremendously helpful in psychological science in general and in mental health in particular. As suggested in the
result, the number of personal pronouns embedded in an essay may divulge psychological states, in this case,
depression.
Negative and positive emotions, on the other hand, had different trajectories on their influence on depression.
Words that pertain to positive emotions had shown a negative influence on depression and words associated with
negative emotions positively influenced depression. This is not surprising because most empirical evidence had
already paved an explanation. Happy people have a lower tendency to get depressed than lonely people. What is
unique in this finding, however, is the employed method. Unlike most studies in this area, the present findings
were found using a unique way of understanding and identifying signs of depression. This might imply two
things. First, it is possible to peek into the person’s psychological experiences by analyzing his/her written
language. Second, the findings may be indicative of the effectiveness of LIWC in determining and labeling signs
of depression in written language.
On the other hand, with suicidal ideation, only I-words (personal pronouns) and positive emotion-related
words significantly predict suicidal ideation. Personal writings of individuals with suicidal ideation seemed to
contain more personal pronouns and fewer words that connote positive emotions. A straight forward implication
would be that college students with suicidal thoughts tend to include more personal pronouns in their essays. If
Chung and Pennebaker (2007) were right, the inclusion of I-words in the students’ essay was not intentional.
Rather, it might appear to be a natural process wherein an individual is unaware of. The most important
contribution of this study resides in the fact that the findings of the present investigation were not drawn from
traditional assessment tools unlike the previous studies on depression and suicide ideation in the Philippine
setting (Sta. Maria, Lee, Estanislao, Rodriguez, Wang, & Liu, 2015; Quintos, 2017; Estrada, et al., 2019). This
paper does not discriminate traditional assessment tools on depression and suicidal ideation. But rather to
suggest a new way of understanding and examining psychological problems. Again, the finding of this study
may suggest the importance of personal writing in assessing psychological problems.
5.3 The influence of LIWC summary variables on depression and suicidal ideation
The results suggest that emotional tone and clout were significantly predicted depression. As individuals get
more depressed, they tend to show less emotional tone in their writing. Likewise, as individuals get depressed,
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 41
their writings may contain less sense of authoritativeness. Although the influence of both factors on the variance
of depression was small, the findings may have a significant implication and contribution to the field of mental
health. With the lack of previous studies in this area, this paper could only provide a naive piece of explanation.
As discussed above, the main possible reason as to why personal writings of depressed people tend to indicate
less emotional tone is due to the positive effect of the writing process on the psychological state (Niles et al.,
2014). In effect, decreasing the level of depression associated with essay writing may be the positive effect of
expressive writing.
Lastly, the result suggests that with all the LIWC summary variables, the only emotional tone had shown a
significant link with suicidal ideation. The trajectory of the influence on the suicidal ideation variance might
imply that individuals with a high tendency of having suicidal thoughts tend to have less emotional tone in
personal writing. It could be that; expressive writing would somehow help lessen the participants’ negative
thoughts. Ultimately, writing current emotional feelings could improve well-being and psychological health.
6. Conclusion
Despite the observed weaknesses and limitations, this study conveys important and significant findings.
First, using the LIWC tool, college students with a higher level of depression and suicidal ideation tend to
include more personal pronouns such as “I”, “me”, and “my” in their writing when instructed to write their
current emotional feelings. Secondly, two out of five LIWC traditional dimensions (I-words and negative
emotions) show a significant link with depression but only the I-words dimension was significantly correlated
with suicidal ideation. Third, out of four LIWC summary variables, only clout and emotional tone show a
negative but significant relationship with depression and suicidal ideation. These LIWC domains were also
found to be significant predictors of depression. However, the only emotional tone was found to have a modest
but significant influence on suicidal ideation. Lastly, the length of the essay was significantly correlated with
depression but not with suicidal ideation.
6.1 Limitations and recommendations for future studies
Interpreting the result of this study should involve some caution. Although the results draw tantalizing
conclusions, this paper has statistical and methodological limitations. The first weakness, as you might already
know, stems from the correlational design. Issues in the statistical analysis used may weaken the ability of the
result to provide a stronger proposition. Second, on the methodological aspect, it is worth remembering that the
essay writing and questionnaire administration did not occur simultaneously. The participants were given several
days to complete their writing. Taking this into consideration, college students might have experienced different
psychological states throughout the process. Also, the administration of the questionnaire was during the final
examination week of the semester. Also, most of the participants wrote their essays after the final exam. This
may imply two things. First, since the scales were administered during the exam week, academic pressure and
stress may have played a significant role in students’ psychological state which then affected their reported
depression and suicidal ideation level. Second, if most of the participants wrote their essay after the exam, the
level of stress that was related to depression measured may have somehow sublimed. These factors should be
brought into consideration for future research in this area.
Nonetheless, this study provides a new breed of information not only for the scientific community but also
for mental health professionals. This may be the first endeavor to unravel the signs of depression and suicidal
ideation in the Philippines using text analysis. This paper also suggests that using a text analysis tool can identify
students with psychological turmoil and help them heal.
Lumontod, R. Z., III
42 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
Acknowledgment: The author would like to send a million thanks to the Ethics Review Board for the ethical
guidance that was followed and supported by this inquiry. Furthermore, the author would like to express his
gratitude to all the students who willingly share their time in participating in this study.
7. References
Beck, A. T., Steer, R. A., & Carbin, M. G. (1988). Psychometric properties of the Beck Depression Inventory:
Twenty-five years of evaluation. Clinical Psychology Review, 8(1), 77–100.
https://doi.org/10.1016/0272-7358(88)90050-5
Beck, A.T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression.
Archives of General Psychiatry, 4, 561-571.
Bernard, J. D., Baddeley, J. L., Rodriguez, B. F., & Burke, P. A. (2015). Depression, language, and affect: An
examination of the influence of baseline depression and affect induction on language. Journal of
Language and Social Psychology, 35(3), 317-326. https://doi.org/10.1177/0261927X15589186
Boroditsky, L. (2001). Does language shape thought?: Mandarin and English speakers’ conceptions of time.
Cognitive Psychology, 43(1), 1–22. https://doi.org/10.1006/cogp.2001.0748
Boyd, R. L., & Pennebaker, J. W. (2015). Did Shakespeare write double falsehood? Identifying individuals by
creating psychological signatures with text analysis. Psychological Science, 26(5), 570-582.
https://doi.org/10.1177/0956797614566658
Brockmeyer, T., Zimmermann, J., Kulessa, D., Hautzinger, M., Bents, H., Friederich, H. C., … & Backenstrass,
M. (2015). Me, myself, and I: self-referent word use as an indicator of self-focused attention in relation
to depression and anxiety. Frontiers in Psychology, 6(1564), 1-10.
https://doi.org/10.3389/fpsyg.2015.01564
Castillo, L. G., & Schwartz, S. J. (2013). Introduction to the special issue on college student mental health.
Journal of Clinical Psychological, 69(4), 291–297. https://doi.org/10.1002/jclp.21972
Chen, C. J., Chen, Y. C., Sung, H. C., Hsieh, T. C., Lee, M. S., & Chang, C. Y. (2015). The prevalence and
related factors of depressive symptoms among junior college nursing students: A cross-sectional study.
Journal of Psychiatric and Mental Health Nursing, 22, 590–598. https://doi.org/10.1111/jpm.12252
Chung, C. k., & Pennebaker, J. W. (2007). The Psychological functions of function words. In K. Fiedler (Ed.),
Frontiers of social psychology. Social communication (p. 343–359). Psychology Press.
Cichocka, A., Bilewicz, M., Jost, J. T., Marrouch, N., & Witkowska, M. (2016). On the grammar of politics—or
why conservatives prefer nouns. Political Psychology, 37(6), 799-815.
https://doi.org/10.1111/pops.12327
Coskun, O., Ocalan, A. O., Ocbe, C. B., Semiz, H. O., & Budakoglu, I. (2019). Depression and hopelessness in
pre- clinical medical students. The Clinical Teacher 16, 345–351.
Dahlin, M., Joneborg, N., & Runeson, B. (2005). Stress and depression among medical students: A
cross-sectional study. Medical Education, 39(6), 594–604.
https://doi.org/10.1111/j.1365-2929.2005.02176.x
Delfabbro, P. H., Malvaso, C., Winefield, A. H., & Winefield, H. R. (2015). Socio-demographic, health, and
psychological correlates of suicidality severity in Australian adolescents. Australian Journal of
Psychology, 68(4), 261-269. https://doi.org/10.1111/ajpy.12104
Dunn, M. S., Goodrow, B., Givens, C., & Austin, S. (2008). Substance use behavior and suicide indicators
among rural middle school students. Journal of School Health, 78(1), 26-31.
Ediz, B., Ozcakir, A., & Bilgel, N. (2017). Depression and anxiety among medical students: Examining scores of
the beck depression and anxiety inventory and the depression anxiety and stress scale with student
characteristics. Cogent Psychology, 4(1). https://doi.org/10.1080/23311908.2017.1283829
Estrada, C. A. M., Nonaka, D., Gregorio, E. R., Jr. Leynes, C. R., del Castillo, R. T., Hernandez, P. M. R., … &
Kobayashi, J. (2019). Suicidal ideation, suicidal behaviors, and attitudes towards suicide of adolescents
enrolled in the alternative learning system in Manila, Philippines: A mixed methods study. Tropical
Extracting signs of depression and suicidal ideation from college students’ writing using LIWC
International Journal of Research Studies in Education 43
Medicine and Health, 47(22), 1-18. https://doi.org/10.1186/s41182-019-0149-6
Garlow, S. J., Rosenberg, J., Moore, J. D., Haas, A. P., Koestner, B., Hendin, H., & Nemeroff, C. B. (2008).
Depression, desperation, and suicidal ideation in college students: Results from the American
Foundation for Suicide Prevention College Screening Project at Emory University. Depression and
Anxiety, 25(6), 482–488. https://doi.org/10.1002/da.20321
Horgan, A., Sweeney, J., Behan, L., & McCarthy, G. (2016) Depressive symptoms, college adjustment and peer
support among undergraduate nursing and midwifery students. Journal of Advanced Nursing 72(12),
3081-3092. https://doi.org/10.1111/jan.13074
Jordan, K. N., Sterling, J., Pennebaker, J. W., & Boyd, R. L. (2019). Examining long-term trends in politics and
culture through language of political leaders and cultural institutions. Proceedings of the National
Academy of Sciences, 116(9), 3476–3481.
Kacewicz, E., Pennebaker, J. W., Davis, M., Jeon, M., & Graesser, A. C. (2013). Pronoun use reflects standings
in social hierarchies. Journal of Language and Social Psychology, 33(2), 125-143.
https://doi.org/10.1177/0261927X13502654
Kuo, E. S., Stoep, A. V., Herting, J. R., Grupp, K., & McCauley, E. (2013). How to identify students for
school-based depression intervention: Can school record review be substituted for universal depression
screening? Journal of Child and Adolescent Psychiatric Nursing, 26(1), 42–52.
https://doi.org/10.1111/jcap.12010
Lamis, D. A., Ballard, E. D., May, A. M., & Dvorak, R. D. (2016). Depressive symptoms and suicidal ideation in
college students: The mediating and moderating roles of hopelessness, alcohol problems, and social
support. Journal of Clinical Psychology, 72(9), 919-932. https://doi.org/10.1002/jclp.22295
Maniam, T., Chinna, K., Lim, C. H., Kadir, A. B., Nurashikin, I., Salina, A. A., Mariapun, J. (2013). Suicide
prevention program for at-risk groups: Pointers from an epidemiological study. Preventive Medicine, 57,
S45–S46. https://doi.org/10.1016/j.ypmed.2013.02.022
Montemarano, V., Troister, T., Lambert, C. E., Holden, R. R. (2018). A four-year longitudinal study examining
psychache and suicide ideation in elevated-risk undergraduates: A test of Shneidman's model of suicidal
behavior. Journal of Clinical Psychology, 74(10), 1820-1832. https://doi.org/10.1002/jclp.22639
Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003). Lying words: Predicting deception
from linguistic styles. Personality and Social Psychology Bulletin, 29(5), 665-675.
https://doi.org/10.1177/0146167203029005010
Niles, A. N., Haltom, K. E., Mulvenna, C. M., Lieberman, M. D., & Stanton, A. L. (2014). Effects of expressive
writing on psychological and physical health: The moderating role of emotional expressivity. Anxiety
Stress Coping, 27(1), 1-19.
Nyer, M., Farabaugh, A., Fehling, K., Soskin, D., Holt, D., Papakostas, G. I., … & Mischoulon, D. (2013).
Relationship between sleep disturbance and depression, anxiety, and functioning in college students.
Depression and Anxiety, 30(9), 873–880. https://doi.org/10.1002/da.22064
Pennebaker, J. W., Chung, C. K., Frazee, J., Lavergne, G. M., & Beaver, D. I. (2014). When small words foretell
academic success: The case of college admissions essays. PLoS ONE, 9(12), 1-10.
https://doi.org/10.1371/journal.pone.0115844
Pudrovska, T., & Karraker, A. (2014). Gender, job authority, and depression. Journal of Health and Social
Behavior, 55(4) 424–441. https://doi.org/10.1177/0022146514555223
Puthran, R., Zhang, M. W. B., Tam, W. W., & Ho, R. C. (2016). Prevalence of depression amongst medical
students: A meta-analysis. Medical Education, 50(4), 456–468. https://doi.org/10.1111/medu.12962
Quintos, M. A. M. (2017). Predictors of suicide ideation among the Filipino youth: Findings from a nationally
representative survey of Filipinos aged 15 to 27. Journal of Multidisciplinary Studies, 6(2), 1-26.
Quintos, M. A. M. (2017). Prevalence of suicide ideation and suicide attempts among the Filipino youth and its
relationship with the family unit. Asia Pacific Journal of Multidisciplinary Research, 5(2), 11-23.
Rotenstein, L. S., Ramos, M. A., Torre, M., Segal, J. B., Peluso, M. J., Guille, C., … & Mata, D. A. (2016).
Prevalence of depression, depressive symptoms, and suicidal ideation among medical students: A
Systematic Review and Meta-Analysis. The Journal of the American Medical Association, 316(21),
Lumontod, R. Z., III
44 Consortia Academia Publishing
(A partner of Network of Professional Researchers and Educators)
2214-2236.
Shneidman, E. S. (1993). Suicide as psychache. Journal of Nervous and Mental Disease, 181, 145-147.
Sta. Maria, M. A., Lee, R. B., Estanislao, S., Rodriguez, C., Wang, J., Liu, Y. (2015). A multivariate analysis of
suicide ideation among university students in the Philippines. Asia-Pacific Social Science Review, 15(1),
46-62.
Tam, W., Lo, K., & Pacheco, J. (2018). Prevalence of depressive symptoms among medical students: Overview
of systematic reviews. Medical Education, 53(4), 345-354. https://doi.org/10.1111/medu.13770
Tan, L., Xia, T., & Reece, C. (2016). Social and individual risk factors for suicide ideation among Chinese
children and adolescents: A multilevel analysis. International Journal of Psychology, 53(2), 117-125.
https://doi.org/10.1002/ijop.12273
Teismann, T., & Forkmann, T. (2015). Rumination, entrapment and suicide ideation: A mediational model.
Clinical Psychology and Psychotherapy, 24(1), 226-234. https://doi.org/10.1002/cpp.1999
Troister, T., Davis, M. P., Lowndes, A., & Holden, R. R. (2013). A five month longitudinal study of psychache
and suicide ideation: Replication in general and high risk university students. Suicide and
Life-Threatening Behavior 43(6), 611-620. https://doi.org/10.1111/sltb.12043
Van Spijker, B. A. J., Batterham, P. J., Calear, A. L., Farrer, L., Christensen, H., Reynolds, J., & Kerkhof, A. J.
F. M. (2014). The Suicidal Ideation Attributes Scale (SIDAS): Community-based validation study of
a new scale for the measurement of suicidal ideation. Suicide and Life-Threatening Behavior, 44(4),
408-419. https://doi.org/10.1111/sltb.12084
Vaughn, L. A. (2018). Contents of hopes and duties: A linguistic analysis. Frontiers in Psychology, 9(757), 1-13.
https://doi.org/10.3389/fpsyg.2018.00757
Wong, Y. J., Uhm, S. Y., & Li, P. (2012). Asian Americans’ family cohesion and suicide ideation: Moderating and
mediating effects. American Journal of Orthopsychiatry, 82(3), 309–318.
https://doi.org/10.1111/j.1939-0025.2012.01170.x
... In recent years, much research has been completed to investigate the link between mental health and linguistic use to get novel insights into identifying suicidal thoughts. Previous works on detecting suicidal thoughts made use of language elements from the psychiatric literature, such as LIWC [15], emotion features [16], and suicide notes [17]. However, the fundamental disadvantage of this analysis approach is that it utilizes language-specific strategies that evaluate individual posts in isolation and will not perform well enough when dealing with diverse or enormous quantities of data. ...
Article
Full-text available
Social networks are essential resources to obtain information about people’s opinions and feelings towards various issues as they share their views with their friends and family. Suicidal ideation detection via online social network analysis has emerged as an essential research topic with significant difficulties in the fields of NLP and psychology in recent years. With the proper exploitation of the information in social media, the complicated early symptoms of suicidal ideations can be discovered and hence, it can save many lives. This study offers a comparative analysis of multiple machine learning and deep learning models to identify suicidal thoughts from the social media platform Twitter. The principal purpose of our research is to achieve better model performance than prior research works to recognize early indications with high accuracy and avoid suicide attempts. We applied text pre-processing and feature extraction approaches such as CountVectorizer and word embedding, and trained several machine learning and deep learning models for such a goal. Experiments were conducted on a dataset of 49,178 instances retrieved from live tweets by 18 suicidal and non-suicidal keywords using Python Tweepy API. Our experimental findings reveal that the RF model can achieve the highest classification score among machine learning algorithms, with an accuracy of 93% and an F1 score of 0.92. However, training the deep learning classifiers with word embedding increases the performance of ML models, where the BiLSTM model reaches an accuracy of 93.6% and a 0.93 F1 score.
... In terms of consistency with the prior literature, negative emotions 28,34,35,[40][41][42] , 1 st person singular pronouns 27,28,30,35,36,40,41 , swear words 34,36,37,39,40,43 , and negations 36,37,40,44 were shown to be positively associated with depression severity. While, positive emotions 31,36,40,43,45,46 , 1 st person plural 28,36,41 , 2 nd person 36,40,44 , and 3 rd person pronouns 36,40,41,44,47 have been found to be negatively associated with depression. Article use has been found to be significantly associated with depression severity, although there is inconsistency regarding the direction of the effect 28,36,37,39,40 . ...
Article
Full-text available
Network theory of mental illness posits that causal interactions between symptoms give rise to mental health disorders. Increasing evidence suggests that depression network connectivity may be a risk factor for transitioning and sustaining a depressive state. Here we analysed social media (Twitter) data from 946 participants who retrospectively self-reported the dates of any depressive episodes in the past 12 months and current depressive symptom severity. We construct personalised, within-subject, networks based on depression-related linguistic features. We show an association existed between current depression severity and 8 out of 9 text features examined. Individuals with greater depression severity had higher overall network connectivity between depression-relevant linguistic features than those with lesser severity. We observed within-subject changes in overall network connectivity associated with the dates of a self-reported depressive episode. The connectivity within personalized networks of depression-associated linguistic features may change dynamically with changes in current depression symptoms.
... Major depressive disorder (MDD), as a common mood disorder, is a kind of severe mental disease from which more than 264 million people worldwide suffer (1). It is well known that MDD is a major cause of disability and a significant risk factor for suicide (2)(3)(4)(5), contributing to a growing global burden on the economy and health systems of the society (6). ...
Article
Full-text available
Background: Major depressive disorder (MDD) is a common and intractable mood disorder. Electroconvulsive therapy (ECT) is a common means of brain electrical stimulation for the treatment of MDD, but the neurobiological mechanism of its clinical symptom relief effect is still to be explored. This study aims to explore how ECT plays a role in depression remissions by investigating the changes of static and dynamic brain network characteristics in MDD patients before and after ECT. Methods: Resting-state functional magnetic resonance imaging (fMRI) scans were obtained from nine MDD patients twice before and after a full course of ECT, all of whom responded to ECT as defined by at least a 50% reduction from baseline Hamilton Depression Scale (HAMD) scores. Both static and dynamic characteristics of the functional brain network were compared between the pre- and post-ECT scans for all participants, and the correlations between changes in clinical symptoms and altered network metrics were also investigated. Results: The clustering coefficient and local efficiency in static brain networks were increased significantly, while the global flexibility of dynamic brain networks was decreased significantly after ECT. Several regions of interest (ROIs) that changed significantly at the local level were also identified, which involved regions of the cerebellum, hippocampus as well as frontal and temporal cortices. Although not significant, the decrease of HAMD scores were associated with trends of changed network metrics after ECT. Conclusions: Our results suggest that ECT may alleviate the depressive symptoms of MDD by decreasing the randomness of the brain network as reflected by changes in both static and dynamic network properties and that the temporal gyrus, frontal gyrus, hippocampus, and cerebellar regions may play key roles in such mechanisms. These findings have important implications for our understandings of ECT and depression. However, this study is limited by a relatively small sample size and the results should be confirmed in larger samples.
... Similarly, retrospective state bad mood showed correlations with emotional tone, negative emotion, sadness, and positive emotion in the expected direction in the cheerful writing scenario. These findings are consistent with previous findings demonstrating emotional tone was associated with depression and suicidal ideation (Lumontod, 2020). These findings are consistent with the trait-based literature on LIWC category correlates with bad mood (Lau et al., 2021). ...
Article
The State-Trait Cheerfulness Inventory–State Version (STCI-S) measures three states of cheerfulness, seriousness, and bad mood as the temperamental basis of humor. The present study investigates (1) the development and psychometric validation of a newly developed short version and (2) test-criterion validity with state measures and language use. Study 1 (N = 933) confirmed the three-dimensional structure, and both the short and standard versions demonstrated strong psychometric properties. Study 2 (N = 617) replicated these findings in a separate sample. Study 3 (N = 750) demonstrated expected intercorrelations with self-report state measures (i.e., inspiration, self-esteem, depression, anxiety). Participants were randomized to expressive writing conditions (i.e., writing about a retrospective cheerful, serious, or bad mood scenario). Unacquainted judges’ and participants’ ratings of three states in their respective scenarios converged (.44 < r < .62). Retrospective states cheerfulness, seriousness, and bad mood were associated with linguistic categories (e.g., emotional tone, clout, achievement, insight) identified in the Linguistic Inquiry and Word Count (LIWC) software. The impact of personality states on word usage may only be evident in specific contexts and situations. Presence of distinctive language use in these writing samples further demonstrated test-criterion validity and practicality of the STCI-S18.
Article
Full-text available
Background Globally, suicide is a significant cause of death among adolescents. Previous studies conducted in high-income countries suggest that students in alternative schools are more likely to engage in suicidal behaviors than those in formal schools. This study aimed to document suicidal ideation and behaviors among adolescent learners enrolled in the Alternative Learning System (ALS) in Manila, Philippines. Methods A mixed methods study was conducted in 24 ALS centers in the city of Manila. ALS centers were stratified according to congressional district and selected using probability proportionate sampling. A cross-sectional survey to determine attitudes towards suicide and prevalence of suicidal ideation and behaviors was administered to 171 learners aged 13 to 17 years old. In-depth interviews with 18 teachers and 12 learners were conducted to explore the school psychosocial environment’s role on learners’ suicidal ideation and behaviors. Exploratory factor analysis was used to extract attitude factors. Fisher’s exact test and Student’s t-test were used to identify differences in sociodemographic characteristics and attitudes towards suicide between learners with or without suicidal ideation or behaviors. Qualitative data were analyzed using thematic analysis. Results Non-specific active thoughts were the most common type of lifetime suicidal ideation (40.9%) while passive ideation was the most common in the past month (13.5%). Aborted suicide attempt was the most frequent behavior in both lifetime (16.4%) and in the past month (4.7%). Non-fatal suicide attempt in the past month was 2.3%, reaching 12.9% for the entire lifetime. Age, sex, education, and attitudes towards suicide were significantly associated with suicidal ideation or behavior. Thematic analysis showed five themes: (1) fostering belongingness, (2) securing learners’ safety, (3) teaching philosophy, (4) teacher and learner beliefs towards suicidal behavior, and (5) availability of school-offered and community-based services. Conclusion Suicidal ideation and behaviors are prevalent among adolescent ALS learners. This study also showed a significant difference in attitudes towards suicide and sociodemographic characteristics between learners with and without suicidal ideation behaviors. It also suggests that the school psychosocial environment, through social norms and learner-teacher interactions, can potentially prevent progression of suicidal ideation to behavior, influence help-seeking, and promote mental health among learners.
Article
Full-text available
People in a prevention focus tend to view their goals as duties and obligations, whereas people in a promotion focus tend to view their goals as hopes and aspirations. The current research suggests that people's attention goes to somewhat different experiences when they describe their hopes vs. duties. Two studies randomly assigned participants (N = 953) to describe a hope vs. duty. Specifically, Study 1 asked participants to describe a personal experience of pursuing a hope vs. duty, and Study 2 asked participants to describe a current hope vs. duty they had. I analyzed these descriptions with Linguistic Inquiry and Word Count (LIWC) 2015. Consistent with earlier research on regulatory focus, participants wrote more about positive outcomes when describing hopes and social relationships when describing duties. The current research suggests that the effectiveness of common regulatory focus and regulatory fit manipulations could depend on participants' freedom to choose the experiences they bring to mind when they describe their hopes and duties.
Article
Full-text available
The increasing threat of suicide as a cause of death, especially among the youth cohort, serves as the raison d"être of this study. Suicide as a cause of death among young people is increasingly becoming a problem in the Philippines. Suicide has been observed to be the 9th leading cause of death among those aged 20-24 since 2003, and responsible for approximately one out of every three deaths among Filipinos aged 10-24. This study attempted to identify the individual and community level predictors of suicide ideation among the Filipino youth using a nationally representative sample of 19,728 Filipino youth aged 15 to 27. A theoretical framework borne out of the combination of Robert Agnew"s General Strain Theory and Travis Hirschi"s Social Bonds Theory was tested in the study for its capability to explain the suicide phenomenon. Results showed that suicide ideation is a by-product of the interplay between poor family relationships, nature of peers the youth is affiliated with, engagement in other deviant behaviors, and availability of counselors. The findings of the study make salient the importance of five factors in suicide: (1) the value of strong family relationships especially between parents and child, (2) the importance of regulating the youth"s involvement in peer groups, (3) the danger of deviant and risky lifestyles, (4) the importance of access to guidance counselors, and (5) the danger of having suicidal peers.
Article
Full-text available
The increasing threat of suicide amongst the youth serves as the motivation of this study. This study was done to add to the understanding of suicide in terms of ideation and actual attempts among the youth in the Philippines. In particular, this study attempted to accomplish the following objectives: (1) determine the prevalence of suicide ideation and suicide attempt among the Filipino youth, (2) determine the prevalent methods of suicide among the Filipino youth, (3) determine the prevalent reasons for suicide among the Filipino youth, and (4) determine if there is a significant correlation between the family and suicide ideation and suicide attempts among the Filipino youth. Results showed that roughly one in every ten Filipino youth aged 15 to 27 have thought of ending their life through suicide before, though only around one in every twenty pushes through with an actual attempt. When they do attempt to commit suicide, this is most frequently because of problems in the family and they utilize a violent method of suicide in the form of slashing of wrist(s) or the non-violent method of ingesting poisonous substances. However, overall, the use of violent methods is more prevalent. When it comes to the factors correlated with suicide ideation and suicide attempts, this study found significant relationships between integration in the family and suicide. The study emphasizes the importance of a strong relationship between the youth and the other members of the family to curb the probability of suicide.
Article
Full-text available
To evaluate the psychological well-being of medical students a relatively new self-report questionnaire (the depression, anxiety, and stress scale) was used alongside an established instrument (the beck depression and anxiety inventories) in a group of medical students. A total of 928 medical students (49.4% female; 50.6% male) participated voluntarily into this cross-sectional study. Students completed the Turkish versions of the depression anxiety stress scale (DASS), beck depression, and anxiety inventories together with a questionnaire about their socio-demographic characteristics. Mild and moderate levels of depression were found in 30.5% and severe and extremely severe levels of depression in 8.5% of students. Mild and moderate levels of anxiety and stress were present in 35.8% of the participants. Depression and anxiety were more frequently reported by female students. Depression was more frequent among first-year students, students in a poor economic situation and those who were not satisf...
Article
Full-text available
Importance Medical students are at high risk for depression and suicidal ideation. However, the prevalence estimates of these disorders vary between studies. Objective To estimate the prevalence of depression, depressive symptoms, and suicidal ideation in medical students. Data Sources and Study Selection Systematic search of EMBASE, ERIC, MEDLINE, psycARTICLES, and psycINFO without language restriction for studies on the prevalence of depression, depressive symptoms, or suicidal ideation in medical students published before September 17, 2016. Studies that were published in the peer-reviewed literature and used validated assessment methods were included. Data Extraction and Synthesis Information on study characteristics; prevalence of depression or depressive symptoms and suicidal ideation; and whether students who screened positive for depression sought treatment was extracted independently by 3 investigators. Estimates were pooled using random-effects meta-analysis. Differences by study-level characteristics were estimated using stratified meta-analysis and meta-regression. Main Outcomes and Measures Point or period prevalence of depression, depressive symptoms, or suicidal ideation as assessed by validated questionnaire or structured interview. Results Depression or depressive symptom prevalence data were extracted from 167 cross-sectional studies (n = 116 628) and 16 longitudinal studies (n = 5728) from 43 countries. All but 1 study used self-report instruments. The overall pooled crude prevalence of depression or depressive symptoms was 27.2% (37 933/122 356 individuals; 95% CI, 24.7% to 29.9%, I² = 98.9%). Summary prevalence estimates ranged across assessment modalities from 9.3% to 55.9%. Depressive symptom prevalence remained relatively constant over the period studied (baseline survey year range of 1982-2015; slope, 0.2% increase per year [95% CI, −0.2% to 0.7%]). In the 9 longitudinal studies that assessed depressive symptoms before and during medical school (n = 2432), the median absolute increase in symptoms was 13.5% (range, 0.6% to 35.3%). Prevalence estimates did not significantly differ between studies of only preclinical students and studies of only clinical students (23.7% [95% CI, 19.5% to 28.5%] vs 22.4% [95% CI, 17.6% to 28.2%]; P = .72). The percentage of medical students screening positive for depression who sought psychiatric treatment was 15.7% (110/954 individuals; 95% CI, 10.2% to 23.4%, I² = 70.1%). Suicidal ideation prevalence data were extracted from 24 cross-sectional studies (n = 21 002) from 15 countries. All but 1 study used self-report instruments. The overall pooled crude prevalence of suicidal ideation was 11.1% (2043/21 002 individuals; 95% CI, 9.0% to 13.7%, I² = 95.8%). Summary prevalence estimates ranged across assessment modalities from 7.4% to 24.2%. Conclusions and Relevance In this systematic review, the summary estimate of the prevalence of depression or depressive symptoms among medical students was 27.2% and that of suicidal ideation was 11.1%. Further research is needed to identify strategies for preventing and treating these disorders in this population.
Article
Background: Mental distress seriously affects the quality of life of medical students. Medical students face mental health problems such as depression, anxiety and hopelessness. The pre-clinical years include substantial changes in a student's life, such as moving cities and losing friends, which can complicate an already stressful life, leading to depression and hopelessness. This study aims to determine the levels of depression and hopelessness, and the habits that can affect these levels, among pre-clinical medical students. Methods: A descriptive cross-sectional study was conducted using anonymous questionnaires composed of the Beck Depression Scale (BDS), the Beck Hopelessness Scale (BHS), and questions about sociodemographic and perceived reasons for happiness or unhappiness. We invited pre-clinical medical students to join the study. Particularly during the pre-clinical years, medical students face the adverse effects of changing cities, social circles co-workers RESULTS: A total of 1015 (70.5%) students participated in the study. The prevalence of depression among pre-clinical students was found to be 27.7 per cent and mild levels of hopelessness were found in 54.4 per cent. A meaningful positive correlation was found between the BDS scores and BHS scores of medical students (r = 0.535; p < 0.0001). The BDS scores of students who were ex-smokers, daily consumers of fizzy beverages and energy drinks, and students with chronic diseases were significantly higher (p < 0.05). The BHS scores of the students who were ex-smokers and daily consumers of fizzy drinks, as well as intersex students, had significantly higher scores than the other groups (p < 0.05). Discussion: A healthy social life and effective psychological counselling services are very much needed for mental well-being among medical students in their pre-clinical year.
Article
Context The prevalence of depressive symptoms among medical students has been found to be higher than among other students because of their longer studying time, higher workload and larger financial burden. Despite the availability of reviews examining this, some have focused solely on one country, whereas others used databases containing papers of solely one language; therefore, the results from reviews might not be comprehensive. Against this background, this overview aims to synthesise the results from all the published systematic reviews of depression among medical students, in order to provide a more accurate result. Methods A systematic search was conducted of online databases for published systematic reviews or meta‐analyses examining the prevalence of depressive symptoms among medical students. The findings of individual studies included in these reviews were extracted and then combined with a random‐effects model. Subgroup analysis was conducted by regions. Results A total of 10 studies were selected in this overview, involving 249 primary studies and 162 450 medical students. MEDLINE was the most popular database used in these studies. The overlapping of primary studies in these reviews was appreciably high, except for three studies that focused on specific countries. The overall pooled prevalence was 27.0% (95% CI, 24.7–29.5%). Significant subgroup differences were detected (p < 0.001). The pooled prevalence among studies in the Western Pacific Region was the lowest, 18.9% (95% CI, 11.7–29.0%), whereas that in Africa (40.9%) was the highest (95% CI, 28.8–54.4%). The top five significant factors associated with depressive symptoms were: (i) year of study; (ii) gender; (iii) personal issues; (iv) family relations or issues, and (v) health status. Conclusions Depression affected around a quarter of medical students in general and 40.9% of students in Africa. It is suggested that medical schools and health authorities should introduce preventive measures to curb the high prevalence of depressive symptoms.
Article
1 Objectives Using a 4‐year follow‐up design, this research investigated Shneidman's model of psychache (i.e., intense mental pain/anguish) as the cause of suicide. Operationalizing suicidal manifestations using suicide ideation, we evaluated Shneidman's assertion that psychache is the prominent predictor of suicide ideation and that other suicide‐related psychological variables associate with suicide ideation only through psychache. 2 Method Eighty‐two undergraduates at elevated suicide risk were assessed at baseline and follow‐up with measures of suicide ideation and three psychological predictors: depression, hopelessness, and psychache. 3 Results At baseline, only psychache and neither depression nor hopelessness contributed significant, unique information to statistically predicting suicide ideation. For 4‐year change in suicide ideation, only psychache and neither depression nor hopelessness provided significant, unique information. 4 Conclusions Results provided partial support for Shneidman's contention of the importance of psychache for suicidal behavior and that other psychological factors are only important to suicide insofar as they relate through psychache.
Article
AimThis study aimed to identify levels of depressive symptoms, social and personal college adjustment and peer support among nursing and midwifery students. Background Student mental health is of international concern, particularly among students who are undertaking professional qualifications in health care. DesignCross-sectional design. Methods Data were collected in 2013 using the Centre for Epidemiology Depressive Symptoms Scale, two subscales of the Student Adaptation to College Questionnaire; and a subscale of the Peer Support Evaluation Inventory with 417 students in Ireland. ResultsFindings indicated that 34% of participants experienced depressive symptoms, 20% were poorly personally adjusted and 9% poorly socially adjusted. Most students had good levels of peer support. Statistically significant relationships were found between all key variables. Students in their second year of study had significantly higher rates of depressive symptoms. Participants who reported having poor relationships with their fathers were at higher risk and had more difficulties personally and socially adjusting to university life and study. The alcohol consumption of participants had a statistically significant relationship with depressive symptoms with higher consumption rates having a positive impact on symptoms. Conclusion The mental health of undergraduates undertaking professional healthcare studies needs to be a key research, educational and clinical priority. High rates of adjustment and mental health difficulties, particularly in the second year of the programme need to be examined and more effective interventions developed.