Local political participation in the Netherlands: an analysis of the influence of online media use for political purpose on political participation.

(1)

Na me: Landstra, Car men St udent nu mber:

Dat e: 4^{t h} of Jul y 2019

St udy Pr ogra m: Ma nage ment, Soci et y & Technol ogy Instit uti on: Uni versit y of Twent e, Enschede

Super vi sor: Dr. P. J. Kl ok

Second Super vi sor: S. Donnell y

Versi on: 1

Wor dcount: 15343

(2)

2 Su mmary/ Abstract:

Politi cal parti ci pati on is decli ni ng and t hat poses real t hreat s t o t he democr ati c val ues i n west er n soci eti es because a de mocr acy needs its citi zens t o parti ci pat e. No wadays, much research is conduct ed on t he rel ati onshi p bet ween onli ne medi a and politi cal parti ci pati on i n t he acade mi c areas of political sci ences, poli cy st udi es and co mmuni cati on sci ences. Thi s st udy buil ds upon i nsi ght s fro m ot her schol ars whil e maki ng use of a recent, large dat a set of ar ound 2500 respondent s, represent ati ve for Dut ch citi zens fro m 2018 duri ng t he muni ci pal el ecti ons.

The ai m of t his t hesi s is to ans wer t he questi on whet her t here is a positi ve associ ati on bet ween Onli ne Medi a Use f or Political Purpose and Offli ne Politi cal Parti ci pati on in t he Net herl ands.

In t his t hesi s different forms of onli ne medi a use and parti ci pati on are consi dered, i dentifyi ng bot h acti ve and passi ve for ms, and t he cr uci al i nfluence of Politi cal Int erest is consi dered. It is concl uded t hat t here is i ndeed a si gnifi cant positi ve associ ati on bet ween Onli ne Medi a Use f or Politi cal Purpose and Of fli ne Politi cal Parti cipati on. Also, t he st udy fi nds t hat Politi cal Int erest consi derabl y strengt hens t hi s associ ati on. Ne xt t o t hat, t he st udy provi des evi dence t o cl ai m t hat t he t ype of Of fli ne Politi cal Parti cipati on is si gnifi cantl y mor e i mport ant i n det er mi ni ng t he strengt h of t he associ ati on t han t he t ype of Onli ne Medi a Use f or Politi cal Pur pose.

Ke y words: Onli ne Medi a, Politi cal Parti ci pation, Int ernet, Politi cal Interest, Associ ati on, Regressi on, Net herl ands, Sampl e and Offli ne.

(3)

3

Tabl e of Cont ent s

Su mmar y/ Abstract: . . . 2

1. Introducti on . . . 4

2. Theor y . . . 8

3. Met hods/ Dat a Coll ecti on . . . 13

4. Res ults . . . 30

5. Concl usi on and Discussion . . . 44

Ac kno wl edg ment . . . 48

Ref erences . . . 49

Dat a Appendi x . . . 51

(4)

4 1. I ntroducti on

1. 1. Soci et al and sci entifi c background of t he proble m

De mocracy and bei ng a de mocrati c soci et y i mpl y t hat st at es needs t heir citizens t o parti ci pat e i n t he syst e m. It ai ms directl y at t he core defi niti on of de mocracy whi ch consi sts of t he wor d

‘de mos’ t hat transl at es t o English as citizens. It is t herefore t hat a de mocracy cannot functi on wi t hout its vit al aspect of politi cal parti ci pati on by citi zens. Ho wever, political parti ci pati on no wadays is under t hreat and has been faced wit h a maj or decli ne t hat can be seen i n for exa mpl e decreasi ng nu mbers i n t he t ur nout of el ecti ons ( Bur den, 2000), decreasi ng nu mber of citi zens t hat are a me mber of a politi cal part y ( Biezen & Pogunt ke, 2014) and decreases i n ot her t ypes of politi cal parti ci pati on. Thi s trend is al ar mi ng especi all y a mong young peopl e Bakker and de Vr eese (2011) ar gue. If t his trend conti nues t o decli ne, t his coul d act uall y pose funda ment al t hreat s t o our de mocracy and its values and consequentl y t o soci et y as we kno w it. To avoi d t his fro m happeni ng, it is t heref ore cruci al t o i nvesti gat e pot ential fact ors t hat can i nfl uence t hi s decli ne i n politi cal parti ci pati on.

One of t he pr oposed fact ors t hat can i nfl uence politi cal parti ci pati on is the pot enti al of t he i nt er net. Si nce t he e mer ge of i nt er net, t he devel opme nt s of onli ne medi a have i ncreased rapi dl y.

Wi t hi n a fe w decades t he i nt er net has gr own fro m its first for m i nt o an uni ma gi nabl e co mpl ex net wor k t hat pr ovi des i nnovati ve opport uniti es for t he bot h i nf or mati on and co mmuni cati on technol ogi es. These t echnol ogi es are quit e ne w and are const antl y subj ect t o i nnovati on and change. Due t o t hat, it is by far not yet kno wn what pot enti al t he i nt er net can have i n maj or soci et al issues and i n thi s case, t he decli ne of politi cal parti ci pati on. St at es and t heir gover n ment s have wi dely adopt ed t hese i nf or mat i on and co mmuni cati on technol ogi es i n an att e mpt t o narr ow t he gap bet ween gover n ment al organi sati ons and its citi zens, by for exa mpl e usi ng co mmuni cati on t echnol ogi es t o i nf or m citizens and part y me mbers about t heir pl ans, poi nt s of vi e w and dail y busi ness, and it is also already consi dered t o be an i mport ant el e ment of a successf ul el ecti on ca mpai gn ( Kr ui ke mei er, van Noort, Vli egent hart, & de Vr eese, 2014, p. 903). But how i s t his i nt er net pot enti al t o moti vat e citi zens t o beco me mor e politi call y acti ve? And is it possi bl e t o use onli ne t ool s t o mobilise citi zens for offli ne politi cal acti on?

Muc h research is already conduct ed on t he pot ential effect s of onli ne medi a use on politi cal parti ci pati on. Schol ars have tri ed t o expl ai n t he possi bl e effect by expl ori ng t heori es on f or exa mpl e soci al medi a behavi our, net wor ks, mobilisati on and rei nf orceme nt powers of t he i nt ernet and parti ci pati on. Vari ous schol ars cl aim t hat t here is i ndeed a positi ve rel ati onshi p bet ween bot h onli ne medi a use and parti ci pati on/ engage ment ( Bakker & de Vr eese, 2011;

(5)

5

Bouli anne, 2015; Kr ui kemei er et al., 2014; Qui nt elier & Vi ssers, 2008). What can be i dentifi ed i n t he literat ure is t hat t here is much vari ati on a mong schol ars. Several schol ars focus onl y on t he effect s of soci al medi a, or mor e specifi c onl y one medi u m such as Facebook i nst ead of onli ne medi a use. Next to t hat, so me schol ars also l ook at a mor e li mit ed set of i ndi cat ors t o meas ure t he parti ci pati on vari abl e by for exa mple i ncl udi ng onl y voti ng tur nout and politi cal i nt erest ( Kr ui ke mei er et al., 2014, p. 903) i nst ead of a mor e br oad defi niti on of politi cal parti ci pati on. Vari ous schol ars also i ntroduce ot her vari abl es t hat might pr ovi de mor e expl anati on on t he associati on such as politi cal i nterest ( Kr ui ke mei er et al., 2014, p. 908) and ti me onli ne ( Qui nt eli er & Vissers, 2008, p. 423). Alt hough t here are many schol ars confir mi ng t he positi ve associ ati on, there are also schol ars t hat fi nd a negati ve associ ati on or no correl ati on at all bet ween onli ne medi a use and politi cal partici pati on ( Theocharis & Lo we, 2016).

In a met a-anal ysi s Bouli anne (2015, p. 534) concludes t hat t he dat a of much research suggest s t hat t here is a positi ve rel ati onshi p bet ween social medi a use and parti cipati on i n ci vi c and politi cal life. Also, Bakker and de Vr eese (2011, pp. 460, 462) ar gue t hat usi ng i nt ernet for ne ws and co mmuni cation pur poses is a positi ve predi ct or for parti ci pati on. Ho wever, Kr ui ke mei er et al. (2014, p. 911) found t hat for t he effect of i nt ernet use for politi cal pur pose on politi cal parti ci pati on, t his effect is rat her rei nforci ng and ver y much dependi ng on one’s politi cal i nt erest. They argue t hat t here is an effect, but t he directi on of t his effect is ver y different per t ype of media use and politi cal i nvol ve ment. Thi s is also confir med by Qui nt eli er and Vissers (2008, p. 423), however t hey add i n their research t hat ti me spend onli ne does not pr edi ct parti ci pati on. To contradi ct t hese results that pr ovi de evi dence t o clai m t hat t here is a positi ve associ ati on, Theocharis and Lo we ( 2016, p. 1475) fi nd i n t heir experi ment t hat t ook pl ace i n Gr eece t hat t he rel ati on bet ween medi a use and politi cal i nvolve ment is act uall y negati ve.

Si nce many schol ars have conduct ed research on t he subj ect, it woul d be logi c t o assu me t hat most i nsi ght s are already discovered and di scussed. Ho wever, little i n known on i nt ernet use f or politi cal purpose and it must be not ed t hat t he different schol ars pr ovi de confli cti ng evi dence and cont esti ng concl usi ons. Theref ore, the ai m of t hi s t hesis is t o test for t he effect s of Onli ne Medi a Use For Politi cal Purpose on Offli ne Politi cal Parti ci pation usi ng t he dat a of a recent represent ati ve sampl e fro m t he Dut ch popul ati on.

(6)

6 1. 2. Research Questi on

To f urt her specify t he ai m of t his t hesi s, t he foll owi ng questi on is for mul at ed t hat has t he central focus i n t hi s paper: To what ext ent is t he use of Onli ne Medi a f or Politi cal Purpose aff ecti ng t he Offli ne Politi cal Partici pati on among Dut ch citizens and t o whi ch ext ent is t his i nfl uenced by ot her f act ors?

To f urt her ans wer t hi s questi on, t hree sub- questions are added i n t his t hesi s. The first sub- questi on ai ms t o i dentify whet her Politi cal Int erest might have an i nfl uence on t he rel ati onshi p bet ween Onli ne Medi a Us e f or Politi cal Purpose and Offli ne Politi cal Parti ci pati on. Thi s questi on is for mul at ed i n t he foll owi ng manner: Does Politi cal Int erest i n l ocal and nati onal politi cs i nfl uence t he rel ati onshi p bet ween Onli ne Me di a Use f or Politi cal Purpose and Offli ne Politi cal Parti ci pati on? Ne xt, t o fi nd mor e i nsight on t hi s rel ati onshi p, t wo questi ons are for mul at ed on t he t ype of Onli ne Medi a Use f or Politi cal Purpose and t ype of Offli ne Politi cal Parti ci pati on. First: Do diff erent t ypes of Online Medi a use For Political Purpose have diff erent eff ect s on Offli ne Politi cal Parti ci pati on? and second: Do t he eff ects of Onli ne Medi a Us e f or Politi cal Purpose diff er per t ype of Offline Politi cal Parti ci pati on? If t he research succeeds t o for mul at e an ans wer t o t hese above- menti oned questi ons, it can pr ovi de ne w i nsi ght s and consequent l y possi bl e t ool s t o combat t he decli ne of political parti ci pati on.

Ther ef ore, it is i mport ant t o expl ore and underst and t he rol e of onli ne medi a i n t his matt er.

In t hi s t hesis, t he foll owi ng chapt ers and sequence can be found. First of all, t he next chapt er is concer ned wit h t he t heoreti cal fra me wor k i n whi ch t his t hesi s can be positi oned. In t hi s chapt er, t he core concept s t hat are central i n t hi s thesi s are descri bed, expl ai ned and t heori sed.

Ne xt t o t hat, gui ded and based upon exi sti ng lit eratur e and research, four hypot heses are creat ed t hat are deri ved fro m t heor y and t he research questi ons. In chapt er 3, t he met hodol ogy of t he research will be central. In t hi s chapt er t he choi ce of research desi gn will be di scussed and expl ai ned. The dat a ori gin and coll ecti on, t he operati onalisati on and t he met hods for t he dat a anal ysi s will also be di scussed. In t he operati onalisati on, a det ail ed descri pti on and expl anati on is pr ovi ded on t he constructi on of t he vari abl es t hat are used i n t his t hesi s. In t he dat a anal ysi s, a st epwi se expl anati on on how t he dat a is anal ysed is pr ovi ded. Conti nuing, chapt er 4 will pr ovi ded i nsi ght s on t he out put s t hat are resulti ng from t he hypot heses t esti ng i n SPSS. Her e, an el aborat e struct ured anal ysi s of t he results can be found. Thi s chapt er is struct ured accor di ng t o t he four hypot heses t hat are for mul at ed i n chapter 2. In t he fi nal chapt er, t he result s of t he anal yses will be su mmarised and will be used t o for mul at e sci entifi call y correct ans wers t o t he questi ons t hat are central i n t his t hesi s. Also, t his fi nal chapt er will cont ai n a fe w paragr aphs

(7)

7

on t he li mit ati ons of t he research and reco mme ndati on regar di ng furt her research on t he t opi c.

In additi on t o t his t hesis, a list of references and an appendi x are i ncl uded. The appendi x is struct ured i n accor dance wi t h t he referenci ng i n the t hesi s. The fi gures and t abl es t hat can be found i n t he appendi x are list ed as a. 1. a. 2. b. 1. et c.

(8)

8 2. Theory

In t hi s secti on t he core concept s will be expl ai ned and t heori sed. Consequentl y, a t heoreti cal fra me wor k will be created t o t heori se t he research questi on and its rel evant sub- questi ons.

Fr o m t he t heoreti cal frame wor k, four hypot heses will be i ntroduced t hat wi ll be central i n t hi s t hesi s.

2. 1. Core concepts

2. 1. 1. Onl i ne Me di a Use for Politi cal Purpose

No wadays i n t he Net herlands, most citi zens make use of t he onli ne medi a that are avail abl e. In 2017, 91, 2 % of t he Dut ch i ndi vi dual s was abl e t o make use of t he onli ne me di a (Int er nati onal Tel eco mmuni cati on Uni on, 2017). In t hi s research, t he central questi on i n concer ned wit h Onli ne Medi a Use f or Political Purpose. Thi s concept i mpli es t hat t he use of t he onli ne medi a has a goal i n whi ch t here i n a politi cal i ncenti ve or politi cal questi on at its core. In t he lit erat ure ma ny different di sti nctions and specifi cati ons are made wit hi n t he scope of t his vari abl e ( Bakker & de Vr eese, 2011; Di mitrova, Shehat a, Strö mbäck, & Nor d, 2014; Qui nt eli er &

Vi ssers, 2008) In t hi s research a di sti ncti on is made bet ween Acti ve and Passive for ms of onli ne me di a use. Thi s di sti nction i n oft en used i n t he literat ure on onli ne medi a (Kr ui ke mei er et al., 2014; Dhavan V. Shah, Cho, Evel and, & Kwak, 2005). Dhavan V. Shah et al. (2005) na me t he t wo di mensi ons, Acti ve and Passi ve Onli ne Medi a Us e, ci vi c messagi ng and onli ne i nf or mati on seeki ng. Kr ui ke mei er et al. (2014, p. 906) defi ne acti ve i nt ernet use as i nt eracti ve or t wo- way co mmuni cati ons and passi ve i nt ernet use as i nvol vi ng one- way communi cati on. Thi s concept ualisati on of Kr uike mei er et al. (2014) will be gui di ng i n t his research.

2. 1. 2. Of fli ne Politi cal Parti cipati on

Politi cal parti ci pati on is oft en concept ualised i n ma ny different underst andi ngs ( Bakker & de Vr eese, 2011; Kr ui ke meier et al., 2014; Qui nt elier & Vi ssers, 2008; Theocharis & Lo we, 2016). In general, many schol ars try t o ai m at a concept ualisati on t hat is cl ose t o a defi niti on li ke ‘an acti vit y done by citi zens t o ai m f or politi cal i nfl uence’. Also Offli ne Politi cal Parti ci pati on is di vi ded in t hi s research i n Acti ve Of fli ne Politi cal Parti ci pati on and Passi ve Of fli ne Politi cal Parti ci pati on. Bakker and de Vr eese (2011) i dentify wit hi n di git al parti ci pati on t he di mensions of acti ve and passi ve parti ci pati on i n t heir research. In t his t hesi s we i ncl ude t he noti on that Passi ve Offli ne Political Parti ci pati on is a pr ocess i n whi ch i ndi vi dual s try t o i nfl uence publi c deci si ons by voti ng and si gni ng petiti ons. On t he contrar y, Acti ve Offli ne Politi cal Parti ci pati on will be defi ned as ‘bei ng politi cally acti ve’ wit h t he excepti on of voti ng and si gni ng petiti ons.

(9)

9 2. 2. Caus al Model

On t he subj ect and its rel ati onshi p, many scholars pr opose different expl ai ni ng t heori es:

parti ci pati on t heori es, medi a t heori es, i nt er net t heori es, net wor k t heori es and many mor e. One of t he first t heori es t hat is oft en used t o expl ai n t he rel ati onshi p bet ween Onl i ne Medi a Use f or Politi cal Purpose and Of fli ne Politi cal Parti ci pation is t he mobilisati on t heor y of t he i nt er net.

Ho wever, t here are schol ars t hat can be cat egori sed as havi ng a rat her pessi misti c appr oach t owar ds t he i nt er net and the associ ati on wit h political parti ci pati on. They clai m t hat spendi ng ti me onli ne does affect t he a mount of ti me t he i ndi vi dual can spend on ot her acti viti es, i ncl udi ng engage ment i n politi cal parti ci pati on. The st udy by Theocharis and Lo we ( 2016) i ndeed fi nds evi dence for t his cl ai m, ar gui ng that t here is act uall y a negati ve associ ati on bet ween Facebook use and politi cal parti ci pati on. On t he contrar y, t his t hesi s is focussed on t he br oader perspecti ve of onli ne medi a use and t heref ore focusses on evi dence of t he mobilisati on t heor y, however it is i nt eresti ng t o keep t he pessi misti c appr oach and evi dence i n mi nd.

Kr ui ke mei er et al. (2014, p. 904) expl ai n t he ar gu ment of t his mobilisati on t heor y as t he vari ous sources avail abl e onli ne and l ower costs of accessi ng t heses sources encour ages citi zens t o lear n mor e about politi cs and t hus i ncrease engage ment. Also, Tol bert and Mcneal (2003, p.

175) found t hat respondent s wit h access t o t he int ernet and onli ne el ection ne ws wer e mor e li kel y t o vot e bet ween 1996 and 2000. In a met a-anal ysis of 36 st udi es, Bouli anne (2015, p.

534) concl uded t hat t he dat a of t he st udi es suggest a positi ve rel ati onshi p between soci al medi a use and parti ci pati on i n ci vi c and politi cal life. Hence, t he foll owi ng hypot hesi s 1 (see fi gur e 2. 1.) is for mul at ed: t he mor e Onli ne Medi a Use f or Politi cal Purpose, t he hi gher t he Offli ne Politi cal Parti ci pati on. Ho wever, it must be not ed t hat t here is evi dence that t his causal effect mi ght be t he ot her way around. Boogers and Voerma n ( 2003, p. 25) ar gue in t heir research on politi cal websit es and parti ci pati on t hat ‘visits t o politi cal Web sit es are pri maril y confi ned t o politi call y acti ve groups. ’ They i dentify t hat t he mobilisati on t heor y act uall y does appl y f or young peopl e. Boogers and Voer man ( 2003, p. 25) t hus i dentify t hat t he use of t he onli ne medi a can al so be expl ai ned by the fact t hat an i ndi vi dual is already li nked t o a more politi call y acti ve gr oup. Also di Gennar o and Dutt on (2006, pp. 311, 312) found evi dence t o beli eve t hi s reversed causati on effect ar gui ng t hat ‘i ncreasi ng i nvol veme nt onli ne a mong t hose who are already engaged offli ne’. Even t hough t hi s t hesi s might not pr ovi de an ans wer t o t he questi on whet her t here is a case of reversed causati on, it is i mport ant t o keep it i n t he back of t he mi nd.

(10)

10

Hypot hesi s 1: t he more Onli ne Medi a Use f or Politi cal Purposes, t he hi gher t he Offli ne Politi cal Parti ci pati on.

Fi gure 2. 1. hypot hesi s 1

Furt her mor e, it is expected t hat Politi cal Int erest has an i nfl uence on t he rel ati onshi p of t he pr oposed hypot hesi s 1. Thi s pheno menon is often referred t o as rei nf orce ment t heor y. Thi s t heor y expl ai ns t hat t he positi ve effect is stronger if citizens are already int erest ed i n politi cs t hus measuri ng t he effect of l ower l evel s and hi gher l evel s of political i nt erest ( Kr ui ke mei er et al., 2014, pp. 905, 912). If t his is t he case Bouli anne (2011, p. 148) suggest s t hat t he medi a beco mes an i nf or mati on tool for t hose t hat are already i nt erest ed. Vari ous schol ars t hus contr ol for Politi cal Int erest i n their research, also Bakker and de Vr eese (2011). They expl ai ned t hat Politi cal Int erest can be expect ed t o account for a consi derabl e a mount of t he vari ance i n i nt ernet use ( Bakker & de Vr eese, 2011, p. 459). Bi mber (1999, p. 413) goes furt her i n t o det ail of t he effect s of Politi cal Int erest by expl ai ni ng t hat cont acti ng a politici an vi a t he i nt er net is co mpani oned wit h vari ous barri ers and onl y t hose t hat are i nt erest ed i n politics fi nd it wort hy t o pay t he ‘cost s’ t o overco me t hese barri ers due t o t heir i nt erest. Buil di ng furt her on t hat, Pol at (2005, p. 442) cl ai ms t hat t his i nfl uence of Political Int erest is t he cr uci al assu mpti on when consi deri ng t he mobilisati on possi bilities of t he i nter net, peopl e must be suffici entl y i nt erest ed.

Ther ef ore, t he foll owi ng hypot hesi s, see fi gure 2. 2, is present t o t est for t his pheno menon.

Hypot hesi s 2: if t he political i nt erest is hi gh, we expect a stronger rel ati onshi p bet ween Onli ne Me di a use f or Politi cal Purpose and Offli ne Political Parti ci pati on t han if Politi cal Int erest is l ow.

(11)

11

To go deeper i nt o t he rel ati onshi p bet ween Onli ne Me di a Use f or Politi cal Purpose and Offli ne Politi cal Parti ci pati on, different di mensi ons of t he vari abl es were i dentifi ed: Acti ve and Passi ve for ms. Many schol ars concl ude t hat t he effect s differenti at e greatl y when t aki ng i nt o account t hese different dime nsi ons. Qui nt eli er and Vi ssers (2008, p. 423) argue t hat onl y so me i nt ernet acti viti es are successf ul i n sti mul ati ng political parti ci pati on. Bakker and de Vr eese (2011) found t hat different i nt er net acti viti es relat e differentl y t o t he political parti ci pati on di mensi ons. They for exa mpl e i dentify seven different i nt er net acti viti es: i nt er net ne ws use, ser vi ces, musi c, cl ub/ or gani sati on, e- mail, soci al net wor ki ng and for u m. What t hey i dentifi ed is t hat for exa mpl e onli ne for ms of co mmuni cati on, such as e- mail and foru m use, wer e mor e positi vel y and si gnifi cantl y rel at ed t o traditi onal passi ve parti ci pati on t han acti viti es such as list eni ng t o musi c and soci al net wor ki ng ( Bakker & de Vr eese, 2011, p. 462). Thi s confir ms Bakker and de Vr eese (2011, pp. 462, 463) hypot hesi s t hat mor e i nt eractive for ms of onli ne co mmuni cati on is positi vel y rel at ed t o parti ci pati on. Next t o t hat, Dhanavan V. Shah, Mc Leod, and Yoon ( 2001, p. 491) concl ude t hat usi ng t he i nt er net for exchange of i nf or mati on, so a rat her i nt eracti ve appr oach, is associ at ed wit h higher l evel s of parti ci pati on. Next t o t hat, Qui nt eli er and Vissers (2008) found t hat t he t ype of onli ne acti viti es rather t han ti me spend onli ne is mor e successf ul i n expl ai ni ng politi cal parti ci pati on. To t est whet her t here are i ndeed different effect s, and whet her mor e i nt eracti ve t ypes of i nt er net use predict Offli ne Politi cal Parti ci pati on bett er, t he foll owi ng hypot hesi s is creat ed (see fi gure 2. 3).

Hypot hesi s 3: Acti ve Onli ne Medi a Use f or Politi cal Purpose expl ai ns Offli ne Politi cal Parti ci pati on bett er t han Passi ve Onli ne Medi a Us e f or Politi cal Purpose expl ai ns Offli ne Politi cal Parti ci pati on.

Ne xt t o t he different effect s of Acti ve and Passi ve Onli ne Medi a Use f or Political Purpose, t hi s pheno mena is i n t heor y also present i n t he vari able Offli ne Politi cal Parti cipati on. Bakker and de Vr eese (2011, p. 460) concl ude t hat t he i nt er net is for all for ms of partici pati on a positi ve pr edi ct or, but t here are differences present bet ween different t ypes of political parti ci pati on.

(12)

12

They f or exa mpl e i dentify t hat t he i nt er net is a stronger predi ct or for ne wer for ms of politi cal parti ci pati on, for exa mpl e di git al parti ci pati on ( Bakker & de Vr eese, 2011, p. 463). Bouli anne (2015, p. 534) also i dentifi es different for ms of Politi cal Parti ci pati on and she found t hat for exa mpl e soci al medi a has a mi ni mal i mpact on parti ci pati on i n el ecti ons in contrar y t o ot her t ypes of parti ci pati on, such as prot est parti ci pati on ( Bouli anne, 2015, p. 532) Not many schol ars have i dentifi ed acti ve and passi ve for ms of parti ci pati on. Theref ore, t he t heor y does not pr ovi de enough evi dence t o predi ct whi ch form i s bett er expl ai ned by Onli ne Medi a Use f or Politi cal Purpose. Ho wever, when l ooki ng int o t he eff orts t hat are needed t o engage i n Acti ve Offli ne Politi cal Parti ci pati on and Passi ve Of fli ne Parti ci pati on, it can be i dentifi ed t hat t he barri ers for engagi ng i n t he Acti ve for ms of Offli ne Politi cal Parti ci pation are much hi gher t han when engagi ng i n Passi ve Offli ne Politi cal Parti ci pati on. Consi der for exa mpl e t he difference i n efforts between voti ng and att endi ng a muni ci pal council meeti ng ( Oudej ans, 2018a). The ‘cost s’ for an i ndi vi dual will be much hi gher generall y when attendi ng a muni ci pal council meeti ng due t o for exa mpl e ti me resour ces. It can t heref ore be assu med t hat voti ng takes l ess ti me t han att endi ng a muni ci pal council meeti ng and t heref ore t he barri ers t o engage i n passi ve for ms are l ower t han t o engage i n acti ve for ms. Thus, if i ndi vi duals engage easi er i n Passi ve Offli ne Politi cal Parti ci pati on, t he chance t hat Onli ne Medi a Use f or Politi cal Pur pose expl ai ns t hi s bett er seems ar gu ment ati ve. Theref ore, it is i nt eresti ng to hypot hesi ze t hi s assu mpti on and check it agai nst t he dat a. The follo wi ng hypot hesi s will be central (see fi gur e 2. 4).

Hypot hesi s 4: Onli ne Medi a Use f or Politi cal Pur pose expl ai ns Passi ve Offli ne Politi cal Parti ci pati on bett er t han Onli ne Medi a Use f or Politi cal Purpose expl ai ns Acti ve Offli ne Politi cal Parti ci pati on.

(13)

13 3. Met hods

3. 1. Research Desi gn

The research t hat is t o be conduct ed is hypot heses testi ng research. The hypot heses i dentifi ed i n t he t heor y secti on will be t est ed vi a a cr oss-sectional research desi gn. Dool ey (2009, p. 265) i dentifi es cr oss-secti onal as quantit ati ve research t hat is not subj ect t o mani pul ati on and i s meas ured at one poi nt i n ti me. Thi s i mpli es dat a fro m a represent ati ve sa mpl e of a popul ati on meas ured fro m one poi nt i n ti me will be used t o t est t he hypot heses. Alt hough t he st udy consi sts of t wo measure ment s i n ti me, it is not appr opri at e t o call it an i nt errupti ve time seri es st udy or l ongit udi nal desi gn because t here was no treat ment i n bet ween t he sur veys and t he sur veys di d not measure t he sa me variabl es.

Cr oss-secti onal is appr opri at e because t he desi gn is abl e t o use a l ar ge dat a set of many respondent s. It is a fast and cheap way of gat heri ng dat a for many cases. Also, it all ows t o t est for mor e vari abl es t han just t he dependent and i ndependent vari abl e. Thi s all ows t he research t o check for possi bl e confoundi ng vari abl es i n t he rel ati onshi p bet ween Onli ne Medi a Use f or Politi cal Purpose and Offli ne Politi cal Parti ci pation. The desi gn is li mit ed t o t he anal yses of behavi our at one poi nt in ti me, and t heref ore no causal i nferences and concl usi ons can be dr a wn. Thi s means t hat for t his research, onl y concl usi ons can be drawn wit h regar d t o correl ati on and associ ation bet ween t he vari abl es. Anot her li mit ati on of the research desi gn because it onl y measures at one poi nt i n ti me is t hat t he results can differ when usi ng anot her ti me fra me.

I mport ant t o not e is t hat in t he chosen research desi gn, it shoul d be checked whet her t he pr oper construct is refl ect ed i n the measure ment and t he pr esence of rando m err or shoul d be checked.

Thi s coul d pose real t hreat s t o t he measure ment construct vali dit y. Thi s is al so t he case wit h st atisti cal i nference vali dit y ( Dool ey, 2009, p. 267). The desi gn however li mits t he possi biliti es t o check for i nt er nal validit y, for exa mpl e t he spuri ousness of t he rel ati onshi p.

3. 2. Cas e Sel ecti on

In t hi s research t he case sel ecti on are Dut ch citi zens, t he respondent s t o the sur vey. The case was sel ect ed on t he basi s of a ne w avail abl e dat a set i n t he cont ext of l ocal vot er research. The avail abl e dat a pr ovi des opport unit y for an i nducti ve appr oach t owar ds t he dat a. The cases wer e rando ml y sel ect ed fro m a represent ati ve sa mpl e fro m t he Dut ch popul ati on i n 2018. Thi s made sure t hat t he cases t hat is wor ked wit h i n t his sa mpl e are a pr oper representati on of t he Dut ch popul ati on i n 2018.

(14)

14 3. 3. Dat a Coll ecti on

The hypot heses will be t est ed usi ng t he dat a Lokaal ki ezers Onderzoek 2018. Thi s research was concer ned wit h l ocal voter behavi our and opi ni ons about l ocal poli ci es. The dat a coll ecti on was fi nanci all y funded by Sti chti ng Ki ezers Onderzoek Nederl and ( SKON). The dat a was coll ect ed vi a t wo sur veys. The sur vey consi st ed of a pre measure ment ( 5 – 20 Mar ch 2018) and a post measure ment ( 22 – 27 Mar ch 2018 and 2 – 24 April 2018) ( Oudej ans, 2018a, p. 2).

The sur vey was present ed t o a represent ati ve sampl e of t he Dut ch populati on. The sa mpl e consi st ed of 3392 parti cipant s (pre- measure ment) and 3380 parti ci pant s (post - measure ment).

The response rat e of t he pre- measure ment sur vey was 75 %. The response rat e of t he post- meas ure ment was 80, 0 %. The sa mpl e was deri ved rando ml y fro m t he LI SS panel. The LI SS panel ( Langl opende Int ernet St udi es voor Soci ale wet enschappen) consi sts of ar ound 5000 househol ds spread a mong t he Net herl ands. The househol ds for t his panel are sel ect ed by Cent Redat a and Centraal Bur eau voor St at estiek. The parti ci pant s are payed for t heir parti ci pati on i n t he sur veys (de Bl ok et al., 2018, pp. 60, 61).

3. 4. Operati onalisati on

The vari abl es t hat are needed for t he anal ysi s are descri bed bel ow. The variabl es consi st of t he vari ous it e ms t hat are deri ved fro m sur vey questions. The questi ons ori ginat e from Oudej ans (2018a, pp. 5, 8, 9, 10, 25, 26) and Oudej ans (2018b, p. 8). The ori gi nal Dut ch questi ons and transl ati ons t o English can be found i n t he Appendi x A. For t he hypot heses t hat are menti oned i n t he t heor y secti on, t he foll owi ng vari abl es are construct ed.

1. Acti ve Onli ne Medi a Use for Politi cal Pur pose 2. Passi ve Onli ne Medi a Use for Politi cal Pur pose 3. Onli ne Medi a Use for Political Pur pose

4. Acti ve Offli ne Politi cal Parti ci pati on 5. Passi ve Offli ne Politi cal Parti ci pati on 6. Of fli ne Politi cal Parti ci pati on

7. Politi cal Int erest

For t he first hypot hesi s, vari abl es 1 and 4 are needed. For t he second hypot hesi s, vari abl es 1, 4 and 7 are needed. For t he t hird hypot hesi s, variabl es 2, 3 and 4 are needed. For t he f ourt h hypot hesi s, vari abl es 1, 5 and 6 are needed.

(15)

15

3. 4. 1. Acti ve Onli ne Me di a Use for Politi cal Purposes

The vari abl e Acti ve Online Medi a Use f or Political Purpose shoul d measure t o what ext ent i ndi vi dual s engage i n activiti es t hat concer n a t wo-wa y- co mmuni cati on during t he use of onli ne me di a for political pur pose ( Kr ui ke mei er et al., 2014, p. 906). Thi s i mpli es t hat t he it e ms t hat ma ke up t hi s vari abl e require communi cati on, int eracti on and i mmedi ate response. These acti viti es are sel ect ed by struct urall y anal ysi ng and assessi ng t he cont ent of t he sur vey.

It e m Questi on

V11_4pr e Di d you do one of t he follo wi ng duri ng t he el ecti on ca mpai gn (i n t he recent weeks) ? Di scussed about t he muni ci pal el ecti on wit h ot her vi a soci al medi a (Twi tt er, Facebook, What s App).

V12_1 Ha ve you made use of t he i nt er net, e- mails, apps or soci al medi a (Twi tt er, Facebook, What s App) t o cont act cit y councill ors, al der men or t he mayor i n the past fi ve years?

V12_2 Ha ve you used t he i nt ernet, e- mails, apps or soci al medi a ( Twitt er, Facebook, What s App) t o cont act l ocal ci vil ser vant s about a l ocal issue i n t he past fi ve years?

V12_3 Ha ve you used t he i nt ernet, e- mails, apps or soci al medi a ( Twitt er, Facebook, What s App) t o cont act ot her citi zens and or gani sations about a l ocal issue i n t he past fi ve years?

Tabl e 3. 1: Li st of it e ms Acti ve Onli ne Medi a Use for Politi cal Pur pose

The vari abl e Acti ve Online Medi a Use f or Politi cal Purposes consi sts of four it e ms, t hat each consi st of an acti vit y t hat is concer ned wit h t wo- way- co mmuni cati on (see t abl e 3. 1.). Fr o m t he pr e- measure ment dat abase, t he acti vit y concer ni ng ‘ di scussi on about t he muni ci pal el ecti ons vi a soci al medi a’ was select ed. This acti vit y was sel ect ed because of its i nteracti ve nat ure and its use of soci al medi a, whi ch is a sub-cat egor y of onli ne medi a. Thi s ite m i s asked i n t he cont ext of a ti me span of the recent weeks, whi ch impl y t he el ecti on ca mpai gn f or t he muni ci pal el ecti ons. Furt her, no ot her it e ms wer e sel ect ed fro m t he pre- measure ment dat abase due t o t he fact t hat t hey di d not suffici entl y mat ch wit h t he concept ualisati on of Acti ve Onli ne Medi a Use f or Politi cal Purpose. Fro m t he post- measure ment dat abase, t hree acti viti es wer e sel ect ed t o contri but e t o t he measureme nt of t he vari abl e Active Onli ne Medi a Use f or Politi cal Purpose.

These t hree it e ms are concer ned wit h onli ne media: i nt ernet, e- mails, apps or soci al medi a.

Ne xt t o t hat, t he acti vities regar d maki ng cont act wit h different i ndi vi dual s, such as cit y councill ors, al der men, mayors, l ocal ci vil servant s or ot her citizens. These ite ms consi st of a ti me span of fi ve years, whi ch i mpli es t hat t his also concer ns t he ti me span i n whi ch t he first ite m is l ocat ed.

(16)

16

To construct t he vari abl e Acti ve Onli ne Medi a Use f or Politi cal Purpose an i ndex was creat ed, addi ng up t he it e ms v11_4pr e, v12_1, v12_2 and v12_3. Thi s i mpli ed t hat an i ndi vi dual t hat vot es ‘ Yes’ i n all questi ons t hat are li nked t o t he ite ms, t his i ndi vi dual receives a val ue of 4 i n t he vari abl e Acti ve Onli ne Medi a Use f or Politi cal Pur poses whi ch is l abell ed Most Acti ve Use.

The val ues of t he vari able Acti ve Onli ne Medi a Use f or Politi cal Purposes ranges fro m 0 t o 4.

Val ue 0 meani ng No Use and val ue 4 meani ng Most Acti ve Use. In or der t o constr uct t he general vari abl e Onli ne Me di a Use f or Politi cal Pur poses, it was necessary t o transf or m t he val ues of t he vari abl e t o a scal e from 0 t o 1. Thi s impl i ed t hat all val ues of Acti ve Onli ne Medi a Us e f or Politi cal Purpose are di vi ded by four resulti ng i n t he foll owi ng val ues. 0 = . 000, 1 = . 250, 2 = . 500, 3 = . 750 and 4 = 1. 00. The di stri buti on of vari abl e Acti ve Onli ne Medi a Use f or Politi cal Purpose can be seen i n fi gure 3. 1.¹

Fi gure 3. 1: hist ogra m Acti ve Onli ne Medi a Use for Political Pur pose 3. 4. 2. Passi ve Onli ne Me di a Us e f or Politi cal Purpose

The vari abl e measures t he Passi ve Onli ne Medi a Us e f or Politi cal Purpose as descri bed i n t he concept ualisati on. The defi niti on ai ms f or acti vities t hat i ncl ude onli ne me di a use wit h a politi cal purpose t hat are charact erised by one-way communi cati on. This i mpli es t hat t he

1 For det ail ed out put s on t he vari abl e constructi on see appendi x B. 1.

(17)

17

acti viti es are not i nt eractive and t hat t here is no i mme di at e response. Acti vities t hat are l abell ed as Passi ve Onli ne Medi a Us e f or Politi cal Purpose are fore most acti viti es in whi ch t he i ndi vi dual recei ves i nf ormati on wit hout engagi ng i n a conversati on.

Tabl e 3. 2: Li st ite ms Passi ve Onli ne medi a Use for Politi cal Pur pose

The vari abl e consi sts of si x it e ms (see t abl e 3. 2.). These it e ms each concern an acti vit y t hat is charact erised as one- way co mmuni cati on. Fr o m the pre- measure ment four ite ms are sel ect ed.

These it e ms concer n activiti es t hat are conduct ed i n t he cont ext of fi ndi ng i nf or mati on about t he muni ci pal el ecti ons. Acti viti es t hat are i ncl uded are visit ed t he websit e of one or more l ocal parti es, visit ed t he website of t he muni ci palit y, made use of a voti ng gui de and read about t he muni ci pal el ecti ons on soci al medi a. The val ues for t hese it e ms range from 0 t o 1. Fr o m t he post- measure ment, t wo ite ms are sel ect ed. The first ite m concer ns t he use of onli ne medi a f or si gni ng a petiti on or l ocal i niti ati ve and t he second it e ms questi ons whet her t he respondent f oll ows l ocal politici ans on soci al medi a. Agai n, in t hese acti viti es it can be consi dered as one- wa y co mmuni cati on. For t he it e m regar di ng t he f oll ow l ocal politi ci ans on soci al medi a, it can be debat ed whet her t he politi cal purpose is cl earl y present. Ho wever, for t his research it is deci ded t hat engagi ng i n thi s acti vit y does pr ovi de the respondent wit h i nf ormati on on t he l ocal politi cs t hat can be put t o pr acti ce for politi cal purpose. It e m v27v1 and v27v2, concer ni ng t he use of onli ne medi a for t he pur pose of i nf or mati on and use of muni ci pal ser vices and facilities, is consci ousl y excl uded. Thi s choi ce was made since t he cont ent of t hese ite ms do not fit t he

It e m Questi on

V10_1pr e To search for i nf or mati on about t he muni ci pal el ecti ons, have you ever done one of t he foll owi ng i n t he l ast fe w weeks? Visit ed t he websit e of one or mor e l ocal parti es.

V10_2pr e To search for i nf or mati on about t he muni ci pal el ecti ons, have you ever done one of t he foll owi ng i n t he l ast fe w weeks? Visit ed t he websit e of t he muni ci palit y.

V10_3pr e To search for i nf or mati on about t he muni ci pal el ecti ons, have you ever done one of t he foll owi ng i n t he last fe w weeks? Made use of a voti ng gui de for t he muni ci pal el ecti ons.

V10_4pr e To search for i nf or mati on about t he muni ci pal el ecti ons, have you ever done one of t he foll owi ng i n t he last fe w weeks? Read about t he muni ci pal el ecti ons on soci al medi a ( Twitt er, Facebook, bl ogs, What s App).

V12_4 Ha ve you used t he i nt ernet, e- mails, apps or social medi a ( Twitt er, Facebook, What s App) t o si gn an i nitiati ve or petiti on on a l ocal issue i n t he past fi ve years?

V36 Do you f oll ow politi ci ans fro m your muni ci palit y on soci al medi a such as Facebook, Twitt er, or Instagra m?

(18)

18

concept ualisati on of t he concept Passi ve Onli ne Medi a Use f or Political Purpose. It is det er mi ned t hat t hese acti viti es l ack Politi cal Purpose.

The vari abl e Passi ve Onl i ne Medi a Use f or Political Purpose is construct ed by addi ng up t he i ndi vi dual it e ms. In or der t o do t hi s, all ite ms t hat are i ncl uded shoul d use same val ues for t heir meas ures. Theref ore, t he v36 fro m t he post- measure ment dat abase had t o be recoded. Aft er t he recodi ng, all ite ms re warded 0 for ans wer ‘ No’ and 1 for ans wer ‘ Yes’. To co mput e t he vari abl e Passi ve Onli ne Medi a Use f or Politi cal Purpose t he it e ms v10_1pr e, v10_2pr e, v10_3pr e, v10_4pr e, v12_4 and v36Rec wer e added up and t his result ed i n a ne w vari abl e t hat had val ues rangi ng fro m 0 t o 6, i n whi ch 0 means ‘ No Use’ and 6 means ‘ Most Active Use’. In or der t o construct t he general variabl e Onli ne Medi a Use f or Politi cal Purpose, it was necessar y t o transf or m t he val ues of t he vari abl e t o a scal e from 0 t o 1. Thi s i mpli ed t hat all val ues of Passi ve Onli ne Medi a Use f or Political Purpose are di vi ded by si x, resulti ng i n t he foll owi ng val ues. 0

= . 000, 1 = . 167, 2 = . 334, 3 = . 500, 4 = . 667, 5 = . 834 and 6 = 1. 00. The distri buti on of t he vari abl e can be found i n fi gure 3. 2.²

Fi gure 3. 2: hist ogra m Passi ve Onli ne Medi a Use for Political Pur pose

(19)

19 3. 4. 3. Onl i ne Me di a Use for Politi cal Purpose

The vari abl e Onli ne Medi a Use f or Politi cal Purpose is concept ualised in chapt er t wo. Thi s vari abl e ought t o measure t o what ext ent i ndi vidual s are usi ng onli ne me di a for politi cal pur poses. Thi s is measured by a list of acti viti es that are concer ned wit h Onli ne Medi a Use f or Politi cal Purpose. To select t he acti viti es t hat are rel evant t o t his vari abl e, t he cont ent of t he sur vey was struct urall y anal ysed and assessed. After t hat, it was deci ded whet her t he sel ect ed ite ms wer e concer ned with one- way co mmuni cati on or t wo- way co mmuni cati on ( Kr ui ke mei er et al., 2014) and consequentl y coul d be consi dered an Acti ve for m or a Passi ve for m of Medi a Us e f or Politi cal Purpose.

The vari abl e Onli ne Medi a Use f or Politi cal Purpose consi sts of bot h Acti ve and Passi ve Onli ne Medi a Use f or Political Purpose. The constructi on of t he vari abl e Acti ve Onli ne Medi a Us e f or Politi cal Purpose can be found i n secti on 3. 4. 1. of t his chapt er. The constructi on of t he vari abl e Passi ve Onli ne Me di a Use f or Politi cal Purposes can be found i n secti on 3. 4. 2. of t hi s chapt er. For t he constructi on of t he vari abl e Onli ne Medi a Use f or Politi cal Pur pose, t he val ues of bot h t he Acti ve and Passi ve vari abl e were alt ered t o a scal e fro m 0 t o 1. Thi s was done t o ma ke sure t hat Passi ve Onli ne Medi a Use f or Political Purpose woul d not have a hi gher wei ght i n t he general vari abl e t han t he acti ve for m, si nce the passi ve for m consi st ed of mor e it e ms. To construct t he general variabl e on Onli ne Medi a Us e f or Politi cal Purpose t he val ues of t he acti ve and passi ve vari able added i n a scal e. Thi s result ed i n a vari abl e ( N=2314) wit h val ues rangi ng from 0 t o 2, i n whi ch 0 is labelled as ‘ No Use’ and 2 as ‘ Most Active Use’. What can be i dentifi ed is t hat mor e t han half of t he i ndi vi duals do not engage i n any of the t est ed acti viti es t hat rel at e t o Onli ne Medi a Use f or Politi cal Purpose. The di stri buti on of the vari abl e can be found i n fi gure 3. 3.³

(20)

20

Fi gure 3. 3: hist ogra m Onli ne Medi a Use for Political Pur pose 3. 4. 4. Acti ve Offli ne Politi cal Parti ci pati on

The vari abl e Acti ve Offline Politi cal Parti ci pati on shoul d measure t o what ext ent i ndi vi dual s engage i n acti viti es t hat concer n politi cal parti cipati on. Acti ve Offli ne Political Parti ci pati on can t hus be i dentifi ed as bei ng politi call y acti ve offli ne, offli ne acti viti es to i nfl uence politi cs, wi t h t he excepti on of si gni ng petiti ons and voti ng ( Bakker & de Vr eese, 2011). Thi s i mpli es t hat t he it e ms t hat make up t hi s vari abl e requir e acti viti es t hat are engagi ng i ndi vi dual s offli ne t o i nfl uence politi cs. These acti viti es are sel ect ed by struct urall y anal ysi ng and assessi ng t he cont ent of t he sur vey.

It e m Questi ons

V11_2pr e Di d you do one of t he foll owi ng duri ng t he el ecti on ca mpai gn (i n t he recent weeks)? Support ed a candi dat e or part y.

V11_3pr e Di d you do one of t he foll owi ng duri ng t he el ecti on ca mpai gn (i n t he recent weeks)? Att ended a ca mpai gn meeti ng or electi on debat e.

V11_1 Whi ch of t he foll owi ng wa ys have you used i n t he l ast 5 years? Cont act (vi a an appoi nt ment, i nt ervi e w or l ett er) wit h a city council or,

al der man, mayor or ci vil ser vant.

V11_2 Whi ch of t he foll owi ng wa ys have you used i n t he l ast 5 years?

At t ended muni ci pal council meeti ng.

V11_3 Whi ch of t he foll owi ng wa ys have you used i n t he l ast 5 years? Visit ed publi c consult ati on eveni ng (s) of your muni ci palit y.

Me mbershi p i n a politi cal part y.

(21)

21

V11_5 Whi ch of t he foll owi ng wa ys have you used i n t he l ast 5 years? Acti ve i n a l ocal acti on gr oup.

Cont act ed a political party i n your muni ci palit y.

Tabl e 3. 3: list of ite ms Acti ve Offli ne Politi cal Parti ci pati on

The vari abl e Acti ve Offline Politi cal Parti ci pati on consi sts of ei ght ite ms, t hat each consi st of an acti vit y t hat is concerned wit h Acti ve Offli ne Politi cal Parti ci pati on (see t abl e 3. 3. ). Fr o m t he pre- measure ment database, t he acti viti es concer ni ng ‘di d you support a candi dat e or part y duri ng t he el ecti on peri od’ and ‘att end a campai gn meeti ng or an el ecti on debat e duri ng t he el ecti on peri od’ were select ed. These acti vities were sel ect ed because of their offli ne nat ure and t heir politi cal ai m. These it e ms wer e asked i n t he cont ext of a ti me span of t he recent weeks, whi ch i mpl y t he electi on ca mpai gn for t he muni ci pal el ecti ons. Further, no ot her it e ms wer e sel ect ed fro m t he pre- measure ment dat abase due t o t he fact t hat t hey di d not suffi ci entl y mat ch wit h t he concept ualisati on of Acti ve Offli ne Politi cal Parti ci pation. Fr o m t he post- meas ure ment dat abase, si x acti viti es were sel ect ed t o contri but e t o t he measure ment of t he vari abl e. These si x it e ms are concer ned wit h acti vities t hat i ndi vi dual s engaged i n duri ng t he past fi ve years t o i nfl uence politi cs. The acti vities regar d maki ng contact wit h different i ndi vi dual s, such as cit y councill ors, al der men, mayors, l ocal ci vil servant s or ot her citizens, t he acti viti es also i ncl ude visits t o meeti ngs and for exa mpl e me mbershi p.

To construct t he vari abl e Acti ve Offli ne Politi cal Parti ci pati on an i ndex was creat ed, addi ng up t he it e ms v11_2pr e, v11_3pr e, v11_1, v11_2, v11_3, v11_4, v11_5 and v11_8. Thi s i mpli ed t hat an i ndi vi dual t hat votes ‘ Yes’ i n all questi ons that are li nked t o t he items, t his i ndi vi dual recei ves a val ue of 8 i n t he vari abl e Acti ve Offline Politi cal Parti ci pati on whi ch is l abell ed Most Politi call y Acti ve. The val ues of t he vari abl e Acti ve Offli ne Politi cal Parti ci pati on ranges from 0 t o 8. Val ue 0 meani ng Not Politi call y Acti ve and val ue 8 meani ng Most Politi call y Acti ve. In or der t o construct t he general vari abl e Offli ne Politi cal Parti ci pati on, it was necessar y t o transf or m the val ues of t he vari abl e t o a scal e fro m 0 t o 1. Thi s i mpli ed t hat all val ues of Acti ve Offli ne Politi cal Parti ci pati on are di vi ded by ei ght resulting i n t he f oll owi ng val ues. 0 = . 000, 1 = . 125, 2 = . 250, 3 = . 375, 4 = . 500, 5 = . 625, 6 = . 750, 7 = . 875 and 8 = 1. 00. ⁴ The di stri buti on of t he vari abl e Acti ve Offline Politi cal Parti ci pati on can be i dentified i n t he fi gure bel ow (see fi gure 3. 4).

(22)

22

Fi gure 3. 4: hist ogra m Acti ve Offli ne Politi cal Parti ci pati on 3. 4. 5. Passi ve Offli ne Politi cal Parti ci pati on

The vari abl e measures t he Passi ve Offli ne Politi cal Parti ci pati on as descri bed i n t he concept ualisati on (chapt er 2). The defi niti on ai ms for acti viti es t hat i ncl ude offli ne acti viti es wi t h a Politi cal Purpose t hat are charact erised by eit her voti ng or si gni ng petiti ons. Thi s i mpli es t hat t he acti viti es do not i ncl ude anyt hi ng el se t han voti ng or si gni ng petiti ons.

It e m Questi ons

V1 Di d you vot e duri ng t he muni ci pal el ecti ons?

V4 At t he sa me ti me as t he muni ci pal el ecti ons of 21 Mar ch, a nati onal referendu m was al so held on t he Int elli gence and Securit y Ser vi ces Act ( Wi v). Di d you vot e i n t his referendu m?

V7 Di d you vot e i n t he parlia ment ar y el ecti ons of 2017?

V11_6 No w we ask you a fe w questi ons about your i nvol ve ment i n l ocal politi cal acti viti es. There are different ways t o raise a political issue or t o i nfl uence l ocal politici ans or t he muni ci palit y. Whi ch of t he foll owi ng ways have you used i n t he l ast 5 years? Si gned a petiti on about a l ocal issue (on paper).

Tabl e 3. 4: list of ite ms Passi ve Offli ne Politi cal Parti ci pati on

The vari abl e consi sts of four it e ms (see t abl e 3. 4.). Fr o m t he pre- measure ment no it e ms wer e sel ect ed. The pre- measure ment does i ncl ude questi ons on voti ng, however t he it e ms questi on whet her i ndi vi dual s are ‘pl anni ng t o vot e i n t he muni ci pal el ecti ons. Therefore, it was deci ded

(23)

23

t o not i ncl ude t he m because t he vari abl e Passi ve Offli ne Politi cal Parti ci pation shoul d meas ur e act ual parti ci pati on. Fr om t he post- measure ment, four it e ms are sel ect ed. The it e ms concer n voti ng i n t he muni ci pal electi ons, voti ng i n t he referendu m, voti ng i n t he parli ament el ecti ons and si gni ng a petiti on on paper i n t he past five years.

The vari abl e Passi ve Offli ne Politi cal Parti ci pati on is construct ed by addi ng up t he i ndi vi dual ite ms. In or der t o do t hi s, all ite ms t hat are i ncl uded shoul d use sa me val ues for t heir measures.

Ther ef ore, t he it e ms v1, v4 and v7 had t o be recoded. In t he sur vey, i ndi vidual s coul d i ndi cat e if t hey vot ed, if t hey di d not kno w, if t hey wer e not all owed t o vot e or if t hey wer e not willi ng t o t ell. Aft er t he recodi ng, all ite ms re war ded 0 for ans wer ‘ No’ and 1 for ans wer ‘ Yes’. The val ues t hat cont ai ned not willi ng t o say, not all owed t o vot e and don’t re me mber wer e coded as missi ng. To co mput e t he vari abl e Passi ve Offli ne Politi cal Partici pati on t he it e ms v1Recoded, v4Recoded, v7Recoded and v11_6 wer e added up and t hi s result ed i n a ne w vari abl e t hat had val ues rangi ng fro m 0 t o 4, i n whi ch 0 means ‘ Not Politicall y Acti ve’ and 4 means ‘ Most Politically Acti ve’. In order t o construct t he general variable Offli ne Political Parti ci pati on, it was necessar y t o recode t he val ues of t he vari abl e t o a scale fro m 0 t o 1. Thi s i mpli ed t hat all val ues of Passi ve Offli ne Politi cal Parti ci pati on are di vi ded by four, resulti ng i n t he foll owi ng val ues. 0 = . 000, 1 = . 250, 2 = . 500, 3 = . 750 and 4 = 1. 00. ⁵ The distri buti on of Passi ve Offli ne Politi cal Parti ci pati on can be found i n fi gure 3. 5.

(24)

24

Fi gure 3. 5: hist ogra m Passi ve Offli ne Politi cal Parti ci pati on 3. 4. 6. Of fli ne Politi cal Parti cipati on

The vari abl e Offli ne Pol itical Parti ci pati on is concept ualised i n t he previous chapt er. Thi s vari abl e ought t o measure t o what ext ent i ndi vi dual s are engagi ng i n offli ne acti viti es t hat are cl assifi ed as Politi cal Parti ci pati on. Thi s is measur ed by a list of acti viti es t hat are concer ned wi t h Offli ne Politi cal Parti ci pati on. To sel ect t he acti viti es t hat are rel evant t o t his vari abl e, t he cont ent of t he sur vey was struct urall y anal ysed and assessed. Aft er that, it was deci ded whet her t he sel ect ed it ems wer e concer ned with Acti ve Offli ne Politi cal Parti ci pati on or Passi ve Offli ne Politi cal Parti ci pati on ( Bakker & de Vr eese, 2011) and consequentl y coul d be consi dered an acti ve form or a passi ve for m of Politi cal Parti ci pati on. The vari abl e Offli ne Politi cal Parti ci pati on consi sts of bot h Passi ve and Acti ve Offli ne Politi cal Parti ci pati on The constructi on of t he vari able Acti ve Offli ne Politi cal Parti ci pati on can be found i n secti on 3. 4. 4.

of t hi s chapt er. The constructi on of t he vari abl e Passi ve Offli ne Politi cal Parti ci pati on can be found i n secti on 3. 4. 5. of t his chapt er.

For t he constructi on of t he vari abl e Offli ne Politi cal Parti ci pati on, t he val ues of bot h t he Acti ve and Passi ve vari abl e were alt ered t o a scal e fro m 0 t o 1. Thi s was done t o make sure t hat Acti ve

(25)

25

Of fli ne Politi cal Parti ci pati on woul d not have a higher wei ght i n t he general vari abl e t han t he passi ve for m, si nce t he acti ve for m consi st ed of mor e it e ms. To construct the general vari abl e on Offli ne Politi cal Partici pati on t he val ues of the Acti ve and Passi ve vari abl e added i n an scal e. Thi s result ed i n a vari abl e ( N=2054) wit h val ues rangi ng fro m 0 to 2, i n whi ch 0 is labelled as ‘ Not Politically Active’ and 2 as ‘ Most Politically Acti ve’. A ne w vari abl e was co mput ed. Thi s vari abl e ranged fro m 0 t o 2.⁶ The di stri buti on of t he vari able Offli ne Political Parti ci pati on can be found i n fi gure 3. 6.

Fi gure 3. 6: hist ogra m Of fli ne Politi cal Parti ci pation

(26)

26 3. 4. 7. Politi cal Interest

The vari abl e Politi cal Interest shoul d measure t o what ext ent i ndi vi dual s consi der t he ms el ves i nt erest ed i n politi cs. Thi s i mpli es t hat t he it e ms that make up t his vari able require questi ons regar di ng i nt erest i n politics. The it e ms for t his vari abl e are sel ect ed by struct urall y anal ysi ng and assessi ng t he cont ent of t he sur vey.

It e m Questi on

V15pr e To what ext ent are you int erest ed i n nati onal politics?

V16pr e To what ext ent are you int erest ed i n l ocal politi cs?

V32a To what ext ent are you int erest ed i n nati onal politics?

V32b To what ext ent are you int erest ed i n l ocal politi cs?

Tabl e 3. 5: list of ite ms Politi cal Int erest

The vari abl e Politi cal Interest consi sts of four ite ms (see t abl e 3. 5.). Fr o m t he pre- measur e ment dat abase, t wo questi ons were sel ect ed regar di ng i nt erest i n l ocal political and i nt erest i n nati onal politi cs. These ite ms wer e asked i n t he cont ext bef ore t he muni cipal el ecti ons. Fr o m t he post- measure ment database, t wo questi ons were sel ect ed t o contri but e to t he measur e ment of t he vari abl e. These t wo it e ms are also concer ned wit h i nt erest i n l ocal politi cs and i nt erest i n nati onal politi cs and were asked aft er t he muni cipal el ecti ons. There are t hus t wo sa me it e ms t hat measure Local Political Int erest and t wo it ems t hat measure Nati onal Politi cal Int erest.

The i ndi vi dual was t heref ore asked bef ore t he muni ci pal el ecti ons and aft er t he muni ci pal el ecti ons t he sa me questions on Politi cal Int erest.

To i ndi cat ed whet her it is a good measure, a factor anal ysi s and a reli abilit y t est is conduct ed (see appendi x C. 1.). The fact or anal ysi s i ndi cat ed t hat onl y one co mponent is extract ed⁷ wit h an ei genval ue of 2. 732⁸. Thi s expl ai ns t hat t he four ite ms t hat are anal ysed t oget her measure 1 co mponent. The Cr onbach’s Al pha val ue is . 845.⁹ Thi s i mpli es t hat t he i nt ernal consist ency of t he i ndex good. Due t o the sa mpl e si ze is of such a l ar ge nu mber N = 2313 it is possi bl e t o wor k wit h t he i ndex and t hus construct t he vari able. The correl ati on bet ween t he it e ms are all positi ve and t here are not maj or differences bet ween t he correl ati ons.^{1 0} The s mall est correl ati on is r =. 449 and t he l ar gest correl ati on is r =. 759. The hi gh correl ati ons can be expl ai ned due t o t he fact t hat i n t he pre-measure ment and i n t he post- measure ment, t he sa me questi ons are

7 See t abl e c. 3. (appendi x C)

1 0 See t abl e c. 5. (appendi x C)

(27)

27

asked. Ho wever, t his correl ati on is not r =1. 00 which means t hat so me i ndivi dual s di d ans wer differentl y i n t he pre- measure ment and i n t he post- measure ment.

Bef ore constructi ng t he vari abl e, t he it e ms had t o be recoded. The it e ms i n t he sur vey had 1 = not i nt erest ed, 2 = fairl y i nt erest ed and 3 = ver y i nt erest ed. In or der t o creat e a scal e t hat meas ures political i nt erest it is i mport ant t o assi gn ‘ not i nt erest ed’ as val ue 0. Theref ore, t he foll owi ng codi ng was creat ed: 0 = not i nt erest ed, 1 = fairl y i nt erest ed and 2 = ver y i nt erest ed.

To construct t he vari abl e Politi cal Int erest a scal e was creat ed, addi ng up t he it e ms v15pr e Rec, v16pr e Rec, v32a Rec and v32bRec. Thi s i mpli ed that an i ndi vi dual t hat votes ‘ not i nt erest ed’

i n all questi ons t hat are linked t o t he it e ms, t his i ndi vi dual recei ves a val ue of 0 i n t he vari abl e Politi cal Int erest whi ch is labell ed ‘ Not Int erest ed’. The val ues of t he vari able Politi cal Int erest ranges from 0 t o 8. Val ue 0 meani ng Not Int erested and val ue 8 meani ng Most Int erest ed. ^{1 1} The di stri buti on of t he vari abl e Politi cal Int erest can be found i n fi gure 3. 7.

Fi gure 3. 7: hist ogra m Political Int erest

1 1 For det ail ed out put s on t he vari abl e constructi on see appendi x B. 7.

(28)

28 3. 5. Dat a Anal ysi s

Fr o m t he pr ocedur e expl ai ned i n t he operati onalisati on^{1 2}, seven wor king vari abl es were construct ed: Onli ne Medi a Use f or Politi cal Purpose, Acti ve Onli ne Medi a Use f or Politi cal Pur pose, Passi ve Onli ne Medi a Use f or Politi cal Purpose, Offli ne Political Parti ci pati on, Acti ve Offli ne Politi cal Parti ci pati on, Passi ve Offli ne Politi cal Parti ci pati on and Politi cal Int erest. These vari abl es are used t o furt her anal yse t he dat a.

Bef ore conducti ng t he anal ysi s, t here are several assu mpti ons t hat needed t o be checked i n or der t o be abl e t o make inf erences for regressi on. The first assu mpti on t hat needed t o be check was t he Li nearit y Assumpti on ( De Veaux, Vell e man, & Bock, 2016, p. 707). Thi s assu mpti on is check vi a t he St rai ght Enough Conditi on and t he Quantit ati ve Dat a Conditi on. The St rai ght Enough Conditi on is checked vi a a scatt er pl ot of resi dual s agai nst t he i ndependent vari abl e or a scatt er pl ot of resi dual s agai nst t he predi ct ed values of t he dependent variabl e. The conditi on is satisfi ed if t he scatt er plot has a hori zont al direction and cont ai ns no patt ern. The Quantit ati ve Dat a Conditi on can be checked by l ooki ng at t he dat aset.

Ne xt t o t hat it is i mport ant t o consi der t he Independence Assumpti on ( De Veaux et al., 2016, p. 708). Thi s will be checked vi a t he represent ati ve nat ure of t he sa mpl e or via a di spl ay of t he regressi on resi dual s t hat is checked for patt er ns, trends or cl u mpi ng. To conti nue, t he Equal Vari ance Assumpti on ( De Veaux et al., 2016, p. 708) needs t o be checked. Thi s was done vi a t he Does t he Pl ot Thi cken? Conditi on, whi ch is a vi sual check of a scatt er pl ot of t he dependent vari abl e agai nst t he i ndependent vari abl e. Also, thi s can be checked for via a scatt er pl ot of t he resi dual s of t he regressi on agai nst t he predi ct ed val ues of t he dependent variabl e.

To concl ude t he assu mpt i ons t hat needed t o be checked, a fi nal check shoul d assess whet her t he Nor mal Popul ati on Assumpti on ( De Veaux et al., 2016, pp. 708, 709) is satisfi ed. Thi s is done vi a t he Nearl y Nor mal Conditi on and t he Outli er Conditi on. The Nearl y Nor mal Conditi on is checked for by l ooki ng at a hist ogra m of t he resi dual s fro m t he regressi on. Due t o t he l ar ge sa mpl e si ze i n thi s research, t his assu mpt i on beco mes of l ess i mport ance.

Hypot hesi s 1: t he more Onli ne Medi a Use f or Politi cal Purposes, t he hi gher t he Offli ne Politi cal Parti ci pati on. To t est for t his rel ati onshi p, t he vari abl es Online Medi a Use f or Politi cal Purpose (i ndependent vari abl e) and Of fli ne Politi cal Parti cipati on (dependent vari abl e) were used. Vi a li near regressi on t he out put s for t his hypot hesi s were coll ect ed. Thi s

1 2 Synt ax co mma nds for t he variabl e constructi on can be found F. 1. i n t he appendi x.

(29)

29

all owed f or results on t he si gnifi cance of t he rel ati onshi p and t he i mpact. Bef ore conducti ng t he anal ysi s, t he assu mptions are checked for.

Hypot hesi s 2: if t he Political Int erest is hi gh, we expect a stronger rel ati onshi p bet ween Onli ne Me di a use f or Politi cal Purpose and Offli ne Political Parti ci pati on t han if Politi cal Int erest is l ow. The second hypot hesis is testi ng for an i nt eracti on effect of Politi cal Int erest on t he above test ed rel ati onshi p. Thi s wi ll be done vi a fi ve regressi on anal yses. First, t he vari abl e Politi cal Int erest was recoded (0 = 0, 1 = 1, 2 = 1, 3 = 2, 4 = 2, 5 = 3, 6 = 3, 7 = 4 and 8 = 4). Thi s i mpli ed t hat t he vari abl e politi cali nt erest Rec had a mi ni mu m of 0 and a maxi mu m of 4. Next, t he co mma nd split fil e by Politi cal Int erest Recoded was i mpl e ment ed. Aft er t hat, a li near regressi on anal ysis was run, resulti ng i n fi ve out put model s i n whi ch each model represent ed a regressi on anal ysis based on one of t he val ues of Politi cal Int erest Recoded. Bef ore conducti ng t he anal ysi s, t he assu mptions are checked for.

Hypot hesi s 3: Acti ve Onli ne Medi a Use f or Politi cal Purpose expl ai ns Offli ne Politi cal Parti ci pati on bett er t han Passi ve Onli ne Medi a Us e f or Politi cal Purpose expl ai ns Offli ne Politi cal Parti ci pati on. The t hird hypot hesi s will be t est ed vi a conducti ng two li near regr essi on anal yses of Onli ne Medi a Use f or Politi cal Purposes on Offli ne Politi cal Parti ci pati on. The i ndependent vari abl es i n t his t est are Acti ve Onl i ne Medi a Use f or Political Pur poses and Passi ve Onli ne Medi a Us e f or Politi cal Purposes. Bef ore conducti ng t he anal ysi s, t he assu mpti ons are checked for.

Hypot hesi s 4: Onli ne Medi a Use f or Politi cal Pur pose expl ai ns Passi ve Offli ne Politi cal Parti ci pati on bett er t han Onli ne Medi a Use f or Politi cal Purpose expl ai ns Acti ve Offli ne Politi cal Parti ci pati on. The fourt h hypot hesi s will be t est ed vi a conducti ng t wo li near regressi on anal yses for Onli ne Medi a Use f or Politi cal Purposes on Offli ne Politi cal Parti ci pati on. The dependent vari abl es i n t his t est are Acti ve Offli ne Politi cal Parti ci pati on and Passi ve Offli ne Politi cal Parti ci pati on. Bef ore conducti ng t he anal ysis, the assu mpti ons are checked for.