Internet Addiction Among KNUST School of Medical Sciences and Dentistry (KSMD) Students - A Preliminary Study in Ghana.

The rising rate of internet accessibility has brought with it an attendant challenge of Internet Addiction (IA). This study assessed the level of IA and its relationships to loneliness, self-esteem, and satisfaction with life focusing on students of the Kwame Nkrumah University of Science and Technology School of Medical Sciences (KSMD) Kumasi, Ghana. The cross-sectional survey method was used with 122 students conveniently sampled from KSMD. From the findings, no severe internet addiction was found, however, 9.84 % and 41.80 % scored in the moderate and mild internet addiction ranges respectively. The results also indicated a weak positive correlation between internet addiction and loneliness, self-esteem and a weak negative but highly significant association with satisfaction with life. While severe IA is not yet a serious problem, 9.83 and 41.3% of the study population suffered from moderate and mild IA respectively and stood the risk of progressing to the severe end of the IA continuum. Counseling was therefore recommended for students to master control over their internet usage.

Keywords: Internet addiction, Loneliness, Self-Esteem, Satisfaction with Life.


INTRODUCTION
The internet has become the most recognized and popular channel for communication 2 commerce, social media, information, academic research and entertainment. 3 In recent years internet usage has increased amongst Ghanaians. 4 Reports from the Miniwatts Marketing Group in 2019, indicated that 10,110,000 Ghanaians, representing 35% of the population of 29,150,000 people currently use the internet and this figure represents an increase of two million on the figures recorded in January 2017. 5 Their research further estimated that as at 30 th June 2019, 11,737,818 Ghanaian use the internet. This represents an internet penetration rate of 39.0% in Ghana at a growth rate of 39.026% over its figures obtained in the year 2000.
The use of the internet has however been recognized as a powerful influence on the daily lives of mankind Journal of Education and Learning Technology (JELT)

RELATED WORK
The overall worldwide picture of the incidence of Internet addiction is one of the diverse and complex scenarios, revealing a wide range of rates, from 1% to 25% 19 or up to 38% in some general populations in parts of the world. 20 . Thus, there appears to be inconsistent IA prevalence rates across different cultures across the globe attributable to a lack of standardized criteria and different research approaches in different ethnic cultures across the world. 21 For example, Uddin et al found that scores on Young's Internet Addiction Test (IAT) among Bangladeshi undergraduate put 47.7% males and 44.5% female students (out of 475 students selected from five universities) within the severe Internet addiction range. 22 This was followed by 27.1% males and 33.9% females scoring within moderate range, with a further 20.7% males and 7.7% females scoring within the mild Internet addiction ranges. Also employing descriptiveanalytic correlation with 408 Iranian undergraduates (150 females and 258 males) from Birjand Islamic Azad University carefully chosen through cluster method of sampling and using Cooper Smith Self-Esteem Inventory (CSEI), Beck Depression Inventory (BDI), and Young's Internet Addiction Test (IAT), Bahrainaian and associates found as high as 40.7% scoring in internet addiction range and a significant relationship between depression, low self-esteem and internet addiction. 23 Also in Iran, a recent cross-sectional research by Salarvand et al found 10% internet addiction rate among 160 Iranian students with a significant adverse relation between internet addiction and Rosenberg's self-esteem score as well as with life satisfaction score. 24 A review of over 100 studies by Wallace showed over 12% males and 5% female students in China experienced severe Internet addiction. 25 Using Pawlikowski's s-IAT for Internet addiction designated as "Internet Use Disorder" (IUD), the short version of Kwon's Smartphone Addiction Scale for "Smartphone Use disorder" (SUD), Interpersonal Reactivity Index (IRI) for "empathy" and items from the Socio-Economic Panel, Germany (SOEP) -Questionnaire for "Life satisfaction", in a recent replicative dual cultural research made a noteworthy discovery about IA and its relationships with selected psychological variables. 26 These include: a substantial inverse association between IUD, empathy and life satisfaction; an inverse association among Smartphone use disorder (SUD) in only Chinese samples, and, a positive association between personal distress and SUD in both samples. The authors explained this as a manifestation of a higher stress susceptibility in social interaction which often placed participants at a heightened risk of developing smartphone use disorder (SUD).
In another cross -sectional survey aimed at establishing the relationship of Internet Addiction (I.A) and depression, loneliness and health related lifestyles among 175 Allied Health Sciences students of the University of Peradeniya; a state university in Sri Lanka, also found 28.6% and 12.0% of students respectively scoring within the mild and moderate addiction ranges, with no participant scoring within severe addiction ranges. 27 Moreover they found a significant positive correlation between internet addiction (I.A.) and both depression and loneliness, and a significant negative correlation between internet addiction and health related lifestyle with male participants having higher I.A scores than females.Another cross-sectional study among 149 University of Sultan Zainal Abidin Medical Students in the state of Terengganu in Malaysia, found 44.9% males and 41.4% females scored within mild internet addiction ranges. 28 In an earlier research Kapahi et al, had adopted similar research approach, and had found 29% of 203 Malaysian high school youth were at risk of (IA). 29 Also in the same country, Masud et al administered internet addiction survey questionnaire to 220 respondents from five universities and established that 6.3 % of males and 3.2 % of females were internet abusers. Additional results from Structural Equational Modeling (SEM) revealed that (IA) had a significant negative influence on their academic performance.  On the other end of the globe, a cross cultural factor analytical research by Sebra et al on 1399 Portuguese and Brazilian internet users aged between 14 and 83 years revealed a negative association between internet addiction and Rrosenberg's self-esteem scores. 31 They also found mostly young Brazilian men aged between 14 -25, was the group exhibiting the most elevated levels of internet addiction. Thus, in terms of gender, research evidence has demonstrated more males affected by IA 32 , as male subjects seem to be more involved in computer-generated activities such as gamming, gambling and cybernetic sex. 33 From the above literature the problem is increasingly becoming widespread that the youth in high schools (and universities) are being negatively affected. In West Africa few peered-reviewed articles have been published on this all-important virgin (IA) research area. For example, a cross -sectional study on IA, locus of control (LOC) and psychological distress in a Nigerian undergraduate sample, 34 found that 27.4%, 12.9% and 9.8% of the sample experienced respectively mild, moderate and severe I.A. 35 The authors noted that in terms of LOC, 69.3% and 30.7% were respectively internally and externally oriented and 24% of the sample had 24.2 psychological distress. Furthermore, their study found that the youth in universities seemed to be more susceptible to (IA) compared to other age groups of society and that 16 -24 year respondents were the most addicted to the internet with much more emotional distress than 25 -31 year olds. 36 IA has received an enormous amount of research effort around the world, especially among the youth in Asia but the same cannot be said about West Africa -especially students in Ghana -Kumasi to be precise. Thus, in Ghana IA has received little research effort. A quick google search of "Internet Addiction in Ghana" yielded mainly e-online publications, and a few academic theses in partial fulfillment of degrees but no peer -reviewed articles. This preliminary research has generally therefore been embarked upon to fill the knowledge gap on IA in Ghana. The specific aims of the present preliminary research therefore were: (i) to estimate the level of Internet Addiction (IA) among KSMD students, and (ii) to establish relationships between IA and psychological variables such as, Satisfaction with Life, Selfesteem, Self-actualization, Loneliness etc.

MATERIALS AND METHODS
This investigation employed a cross -sectional online survey method in finding the level of internet addiction among 122 student volunteers conveniently sampled from the entire School of Medical Sciences and Dentistry in the college of health Sciences, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana. Measures used included a questionnaire covering the socio-demographic particulars, Internet Addiction Test (IAT), Loneliness Scale, Index of Self -esteem, Satisfaction with Life and Common Belief for Students, all of which have been described in detail below. An internet link using google form was created for participants to log on and respond after going through the participants' information sheet including assurances of confidentiality and consenting to voluntarily participate in this research. Apart from the IAT, all the measures used in this current were obtained from Corcoran and Fisher's "Measures for Clinical Practice -A Sourcebook" New York, the Free Press, volumes 1 & 2.

INTERNET ADDICTION TEST (IAT)
The Internet Addiction Test (IAT) is a widely used internet screening test developed by the Centre for Internet Addiction, USA. It was constructed by Dr. Kimberly Young, the director, for the measurement of Internet addiction and compulsivity. It consists of 20 questions examining IA symptoms and each answer rated on a five-point Likert scale from 0 -5; respectively, Not applicable, Rarely, Occasionally, Frequently, Often and Always). "To date, Young's Internet addiction test (IAT) is the most commonly used measure for the research into internet addiction" 37 For example, most of the IA research reviewed in this literature used IAT . 38 The authors chose to use IAT since it is the most commonly used and to be able to have a fair comparative idea of IA in Ghanaian youth. In the pretesting before the research the researchers obtained a Cronbach's Alpha value of 0.859 (85.9%) for internet addiction test (IAT) which suggests a high consistency of internal construct.

LONELINESS RATING SCALE (LRS)
Loneliness Rating Scale (LRS) by Scalise and Ginter consists of 40 items divided into two parts namely; frequency (Part A) and intensity (Part B). 39 The LRS possesses an impressive psychometric properties of 0.82 to 0.89 alpha Cronbach value, six week stability test -retest reliabilities ranging from 0.61 to 0.71, 0.65 to 0.70 for frequency and intensity subscales respectively. For each of the two parts, participants are instructed to indicate which sentence best describes their loneliness experience from 4-and 5-part Likert-scales respectively. An example of a typical statement is as follows; "___When I experience loneliness, I feel low (PART A) …The feeling of being low is: ____... (PART B)" Pretesting revealed a Cronbach alpha value of 0.968 (96.8%) and 0.977 (97.7%) for parts A and B respectively. The authors chose LRS for its impressive psychometric properties and to explore whether each of its four subscales namely; depletion, isolation, agitation and dejection could contribute new knowledge. Moreover, its validity figures suggest a moderate concurrent validity with the commonly used UCLA loneliness scale.

INDEX OF SELF -ESTEEM (ISE)
ISE by Hudson consists of 25 questions which assesses self-esteem problems often fundamental to socio -psychological difficulties. 40 Responders were instructed to quantify how they saw themselves on a Likert scale from a 5 item Likert scale of 1 (none of the time) to 5 (Most of the time) on questions such as "… I feel that people would not like me if they really knew me well…" The sourcebook reports an internal stability of a mean of alpha Cronbach of 0.93 and a two -hour test retest correlation of 0.92 41 pretesting revealed 71.6% internal consistency. The authors chose to use this measure over the popular Rosenberg's self-esteem scale because of its cut -off points, its excellent psychometric properties and because of the claim "it has a wide range of utility for a number of clinical problems" by Corcoran and Fisher, 1994 (p. 383).

SATISFACTION WITH LIFE SCALE (SWL)
SWLS consists of 5 -items instrument that assesses a person's own personal biased cognitive -judgmental evaluation of their quality of life. 42 With an impressive stable internal consistency of 0.87 alpha, 0.82 test -retest two -month reliability and with a high correlation with nine measures of subjective wellbeing the SWLS is regarded as one of the best key component of an individual's mental wellbeing. It uses a 7 item Likert scale from "strongly disagree" to strongly agree" to sentences such as "In most cases my life is close to ideal". An Alpha Cronbach value of 0.848 (84.8%) was obtained in a pretesting. SWLS has been used in numerous researches. Moreover, several studies have proven that the SWLS has satisfactory psychometric properties with internal consistency normally between Cronbach's alpha values of 0.77 and 0.89. 43

RESULTS
This study was designed with two main objectives in mind, namely; to estimate the level of IA and to establish relationships between IA and psychological variables of interests listed and described above by using Statistical Package for Social Sciences (SPSS) 20 by IBM Corporation. Before then a descriptive analysis was performed on the demographic characteristics obtained from the participants and the results are as documented below. with first year students participating the most in this study; 41 (36.61%) followed by third years 29 (23.77%) and sixth years with 19 (15.57%). The rest were 4 th years, second (2 nd ) years and fifth (5 th ) years with 16, 8 and 4 respondents each representing 13.11%, 6.56% and 3.28% respectively. Also, in terms of religion, majority 115 of the respondents were Christians with a percentage of 94.26%, followed by 4 Muslims with a percentage of (3%) and 2 (2%) of the respondents were either Buddhists or Atheists as shown in table 1 below.  Table 2 below shows the internet score ranges obtained from the current study participants according to the score range / cut offs prescribed by the constructor of the IAT.   Figure 1 above, 59 of the respondents (48.36%) of the respondents have no problem with internet usage (i.e., normal users), whereas 41.80% are mildly addicted to the internet, implying they spend a little time on the internet usage but generally can control internet usage. Also, a total of 12 respondents (9.84 %) however indicated they spend a moderate amount of time on the internet and at the borderline, but no participant scoring within the severe internet addiction range.

OBJECTIVE 2: RELATIONSHIPS BETWEEN IA AND PSYCHOLOGICAL VARIABLES OF INTEREST
To achieve objective 3, a correlational analysis was performed on the variables to ascertain relationships between internet addiction and psychological variables of interest and shown in table 3 below.

DISCUSSION OF RESULTS
Of the 122 SPs, 59 respondents constituting the majority (48.36%), scored between 0 -30 on the IAT; suggesting a normal internet usage as presented in table 2. A major finding in the current study is that none of our Study Participants (SPs) is severely addicted to Internet use, since no one scored within (80 -100) range on the IAT. This is in line with research findings from Malaysia 44 who did not find any of their participants severely addicted to internet use. However, this is in sharp contrast to research findings of IA rates of 47.7% males and 44.5% female students in Bangladeshi undergraduates 45 and in 12% male and 5% female students in China from a systematic review of over 100 studies on IA by 46 . It is also in contrast with 0.5% in Italian students 47 , 8.2%  undergraduates by. 51 Two factors, namely; majority female participation and the course of study of our participants may have accounted for finding of no respondent scoring within internet addiction range (80 -100). Firstly, since more males are vulnerable to IA than females globally, as male subjects seem to be more involved in computer-generated activities such as gamming, gambling and cybernetic sex etc. 52 It is not surprising that no respondent was found to be severely addicted to internet in this study with more female voluntary participants of 72, representing 59.02% than their male compatriots, with a participation of 48 representing 39.34% of the SPs. Had more males voluntarily participated, perhaps some could have scored in addiction ranges.
The second factor may have been the course of study of the participants. The general attitude towards medical education in Ghana is that, it is academically demanding in line with research evidence that suggests that health professional trainees have reported higher perceived stress than general population and students in other academic fields 53 . It is therefore possible that students in Allied health and medicine will be more resistant to the problematic use of the internet than their counterparts pursuing other university programs. This may be supported by the fact that the two recent researches by Chamika and Diasthat used 175 and 149 health trainees (Allied health and Medical) respectively, did not find any of their SPs scoring within severe IA range. 54 Another 51 respondents constituting (41.80%) of our SPs who scored in the mild IA range spend a little more time on the internet than desired, even though they maintained a bit more control over their online activities with a score range of (scores 31 -49). This finding is just slightly lower than 49% found by Haque et al, in Malaysian medical students and 57.5% in Italian subjects. 55 In contrast, it is higher than 27.4%, mildly found by Nduanya et al 56 in Nigerian undergraduates. Also, the finding of 12 respondents (9.84%) scoring within moderate IA (scores 50 -79) is lower than 12.0% moderate addiction found by Chamika and Dias 57 in Sri Lankan students. It is also lower than 21.9% and 32% of the study participants who scored within moderate who had occasional or frequent difficulties due to Internet usage found by Servido 58 and Haque et al 59 respectively. It is also lower than 12.9% found by Nduanya 60 in Nigeria. As per the discussion above, it is this 9.83% of our study participants that stand the risk of IA if they are not counselled to control their internet use with free internet connectivity recently made available in this university. Internet availability and differences in number of study participants may be major factors influencing the current results under discussion. This may be explained by the fact that internet speed and availability may be much better in most of these countries whose researchers have been reviewed in this study, compared to Ghana. Nigeria for example, has been estimated to have 61.4% internet penetration compared to 39.0% in Ghana according to 2019 figures byMiniwatt Marketing Group. 61 , (Thus, it is not surprising that the current study found no addiction and relatively higher SPs scoring within normal and moderate IA ranges. Also, whereas internet availability in Asian and developed countries are obviously comparatively higher and faster, the same cannot be said about middle income country such as Ghana. Moreover, numerous institutions of higher education and colleges the world over, have long been offering high speed unrestricted internet connectivity to their students 62 which has recently been introduced in KNUST. Since addictive behaviors develop over time, it is not surprising the current finding revealed only 9.3% moderate internet users. Having said that, it is possible per the findings thatsome tertiary students in Ghana may develop IA, if the current internet availability continues for years to come as has happened in neighboring Nigeria and other developed countries. Another factor worth mentioning in discussing the IA picture in relation with most of the reviewed IA researches, is the number of participants. 63 Thus, most of these researches whose participants scored within Internet Addiction ranges (i.e. Between 80 -100, on the IAT), had participants in excess of 159; in fact, 475, 408, 916, 220, 835 and 160 respectively which were all far greater than 122 SPs used in this study.

RELATIONSHIP OF IA WITH PSYCHOLOGICAL VARIABLES OF INTEREST IA AND SATISFACTION WITH LIFE (SWL).
The findings revealed that Internet Addiction Test (IAT) had a weak highly significant negative correlation with Satisfaction with Life (r = -0.237; p = 0.009). This is suggestive of the fact that as individual's internet use increases, his/ her subjective cognitive assessment of general life satisfaction or quality of life decreases and vice versa. This study replicates findings by Lachmann et al 64 , who found that Internet Use Disorder (IUD) was negatively associated with life satisfaction in a dual cultural (Chinese and German) samples, except that the correlation in this study was weaker than they reported. Interestingly, Lachmann 65 also found a negative association between Smart gadgets Use Disorder (SUD) and life satisfaction in their Chinese samples. This current finding on the other hand is in sharp contrast with findings by Erdogan & Yildirim, 66 who found a weak but significant positive relationship between internet addiction and life satisfaction in Turkish vocational students.

IA AND INDEX OF SELF ESTEEM (ISE)
In the literature, low self-esteem seems to be one of the most powerful risk factors for internet addiction 67 . The findings however revealed an insignificant weak positive correlation between IAT and Index of Self Esteem (r = 0.173; p = 0.056). This replicates findings by Bahrainian et al., (2014) 68 who found a significant positive correlation between selfesteem and internet addiction in their participants. It must be added however, that the correlation was a weak one in the present SP. Thus, as the individual's score on the ISE increases (indicating lower self-esteem), internet use and possible addiction goes up at a relatively slower pace. This finding is however, in contrast to those of Seabra et al 69 and Salarvand et al 70 who found a significant but inverse relation between IA and self-esteem scores. Thus, as the individual's score on the self-esteem increases (indicating lower self -esteem), internet use and possible addiction increases. In fact, the possible explanation for the contrast might be their usage of different self-esteem scales (Rosenberg's Self-esteem Scale) whose scoring (described on page 393 of 71 ), is direct opposite of the scoring ISE used in this current study.

IA AND LONELINESS
Loneliness was assessed in two ways namely; its frequency and intensity in this study. It was established that both the frequency and intensity had a weak positive non-significant relationship with IA (frequency; r = 0.051; p = 0.581; intensity; r = 0.087; p = 0.341). This is in agreement with research evidence from Ezoe at al 72 who found a positive relationship between internet addiction and loneliness. Therefore, internet use leads to loneliness which separates internet addicts from real life social interactions. The observed weaker correlation may be as a result of females being the majority participants in current study. There is a possibility of gender differences in possible actions where females will engage in (e.g. seek social contact and/ or intimacy), when lonely, than what males may likely do (e.g. go online for games, cybernetic sex, operate gadgets etc.).

LIMITATION OF THIS STUDY
The findings from this current study ought to be interpreted cautiously because of a number of limitations. Firstly, the findings of this research were based on only 122 female-dominated voluntary participants. The researchers are of the opinion that this low voluntary participation by the students limited the statistical power and that it is possible that there may be students who are already addicted but failed to participate. Widening the scope to cover the whole student population of the university and possibly a few other Universities in the Kumasi metropolis, may help obtain a much better picture of use of internet and its addiction in Ghana.
Secondly, this study used cross-sectional methodology which cannot establish a causal relationship between IA and the psychological variables studied. Thirdly, this research was limited to KSMD students, thus making it very difficult to even generalize its conclusions to KNUST students, and much less to all tertiary students in Ghana.
A comparative research with a bigger representative sample from different colleges and universities in Ghana is very much recommended to obtain a much better picture of internet addiction among the youth of Ghana

CONCLUSION AND RECOMMENDATION
This preliminary research revealed that while severe IA is not yet a serious problem, 9.83% and 41.3% of the study population suffer from moderate and mild IA respectively and, stand the risk of progressing to the severe end of the IA continuum, if not counselled to master control over their internet use. This study also revealed that participants who are excessive internet users are prone to be unsatisfied with life, lonely, have low esteem, and probably tend to derive solace in the use of the internet. However, there was no correlation between IA and common belief among this study's SPs. Findings from this research will add to the extant research evidence on internet use and Addiction in the youth in Ghana and Sub Saharan Africa as a whole.